In 2009, the orthopedic team at the University of Minnesota introduced the concept of fracture mapping and reported the method of two-dimensional (2D) fracture mapping for the first time. The procedure involves the following steps (Figures 4A–C): Initially, 3D reconstruction images of all patients are obtained, either directly from patient examination data or by reconstructing the fracture model using software such as OsiriX (Pixmeo, Bernex, Switzerland) or Mimics (Materialise, Leuven, Belgium). The optimal viewing angle to best display the fracture line is determined for each perspective. Subsequently, in software like Fireworks (Macromedia, San Francisco, California) or Photoshop, the bone marks are aligned through rotation and scaling, and the fracture images are registered with a standard template. The fracture image is then superimposed onto the template, and a fracture line is drawn. Ultimately, a fracture map is generated by overlaying these lines, which directly illustrates the morphological distribution and areas prone to fractures. However, this technique has several limitations: firstly, it does not involve reducing the fracture fragments, which can lead to inaccuracies in describing the fracture line if there is significant displacement or rotation of the fragments; secondly, although it is based on 3D reconstruction of CT data, it only displays the fracture line distribution from a specific 2D perspective, essentially categorizing it as a 2D fracture mapping technique.

In 2013, Cole et al. (2013) enhanced the existing technology for 2D fracture mapping. To more accurately depict the characteristics of intra-articular fractures, they shifted from relying on a specific perspective of 3D reconstruction to utilizing the fracture lines from CT scans taken 3 mm below the articular surface (Figures 4D–G). These images were then overlaid onto a standard template in Adobe Fireworks or Photoshop to create the final 2D fracture map.

The difference between the methods in Section 3.1.1 and Section 3.1.2 is as follows: the method described in Section 3.1.1 uses 3D reconstructed images to create the fracture map, whereas the method in Section 3.1.2 utilizes the original CT cross-sectional images (CT cross-sections 3 mm below the articular surface) for fracture mapping. However, similar to the method in Section 3.1.1, the method in Section 3.1.2 does not reduce the fracture fragments and only shows the distribution of fracture lines in a 2D view.

In 2014, Mellema et al. (2014) further refined the 2D fracture mapping technique. Unlike previous iterations, this method involved reducing the fracture fragments prior to mapping (Figure 5). 1) In the initial phase, the “Paint Effect” and “Threshold Paint” tools in the 3D Slicer software are employed to manually annotate bone structures on axial, sagittal, and coronal CT images Figure 5. This process is followed by the generation of a 3D polygonal mesh reconstruction. 2) Subsequently, the model was imported into Rhinoceros (McNeel, Seattle, WA, United States) for fragments reduction. In addition, Mimics and 3-Matic (Materialise, Leuven, Belgium) are capable of obtaining three-dimensional mesh reconstruction images from the same perspective as the standard template: Initially, the Mimics software is employed to reconstruct the mask of the fracture model, utilizing the “Edit Mask” function to separate the fracture fragments; Subsequently, the fracture model is imported into 3-Matic, where the “Interactive Translate” and “Interactive Rotate” functions are utilized to reposition the fracture and align the standard model with the fracture model. 3) Finally, after acquiring the 3D mesh reconstruction images from the same perspective as the standard template, these images were imported into Fireworks (Macromedia Inc., San Francisco). Following the accurate alignment of the images with the two-dimensional template, the fracture lines were delineated.

In May 2017, Dugarte et al. (2018) introduced the 3D fracture mapping technique for the first time through a controlled study comparing 2D and 3D maps of scapula fractures. The process involves the following steps: First, the fracture model is reconstructed in Mimcs (Figure 6A). Subsequently, the reconstructed fracture model is imported into Geomagic (3D Systems, Rock Hill, SC), where the “crease angle” tool is employed to segment the fracture model into individual fracture fragments (Figure 6B). Thereafter, best-fit algorithm of Geomagic is utilized to assist in identifying the optimal overlapping position, with the contralateral side serving as a template for the reduction of the fracture fragments (Figures 6C,D). Finally, with the aid of the transparent mode, the fracture line is transferred onto the three-dimensional model of the contralateral side (Figure 6E).

In 2017, Xie et al. (2017) introduced a novel method for 3D mapping of Hoffa fractures, elaborating on the methodology in the appendix of their study. Although this method shares similar principles with the approach by Dugarte et al. (2018), the operational techniques differ. The process of the fracture mapping method described by Xie et al. is as follows (Figure 7): First, the mask of fracture model is reconstructed using Mimics, and the “Edit Mask” function is employed to separate the fracture fragments. Subsequently, the fracture model is imported into 3-Matic, where the “Interactive Translate” and “Interactive Rotate” functions are utilized to realign the fracture and standard models. Finally, the “Transparency” of the standard model is adjusted to “High,” and the “Curve” tool is used to delineate the fracture line on the standard model.

Notably, the virtual reassembly of the fragments deviates from the software’s fitting algorithm, relying instead on manual manipulations such as translation and rotation, enhancing both the universality and accuracy of the fracture reduction. A significant advantage of this approach is that both Mimics and 3-Matic are products of Materialise, facilitating seamless software integration and model importation. Furthermore, this method has standardized steps that have set a precedent for subsequent 3D fracture mapping protocols.

In 2020, Turow et al. (2020) reported on the fracture morphology of scaphoid fractures. Their methodology involved several steps: initially, reverse modeling of the fracture model was performed using 3D Slicer; subsequently, fracture reduction was completed using Artec Studio (Autodesk Inc., San Rafael, CA, United States), followed by overlapping and aligning the fracture model with the standard model; finally, a 3D fracture map was created using Rhinoceros. This has become one of the most commonly used methods in subsequent studies on fracture mapping.

In addition to this, a semi-automated method for 3D fracture mapping was reported in 2023 by Mys et al. (2023) Firstly, Matlab R2021b (The Mathworks Ltd., Natick, MA, United States) and C++ were used to identify separated fragments and semi-automatically separate incompletely separated fragments. This workflow is capable of efficiently processing large amounts of CT data. Manual corrections were performed in Amira (Thermo Fisher Scientific, Waltham, MA, United States). The boundary line of each fracture fragment was then identified semi-automatically. The fracture fragments were then reset. The fracture line of each restored fragment was automatically projected onto the standard model by nearest point matching. The semi-automated mapping of fractures significantly enhances the efficiency of the mapping process.

First, the image of the fracture map is imported into MATLAB, and the ginput command is utilized to convert the fracture lines into (x, y) coordinates. Subsequently, the point coordinates are exported to Excel and re-imported into MATLAB. Based on these coordinates, the data density is calculated by executing a MATLAB script file, with the results displayed in the form of a heatmap. The density of each point is computed by summing the weighted inverse square distances of other points (Mellema et al., 2014; Hadad et al., 2018). Additionally, it has been reported by some scholars that the external lighting function in Photoshop can be utilized to create a heat map as well.

The principle underpinning 3D fracture heat mapping is akin to that of 2D fracture heat mapping, the key distinction being that the former necessitates 3D spatial coordinates. Zeng et al. (2022) reported a method for 3D fracture heat mapping. After delineating the fracture lines in 3-Matic, these lines were subsequently imported into AutoCAD (Autodesk Inc., San Rafael, CA, United States). The fracture line data was then extracted within the AutoCAD software, with a consistent spacing of 0.1 mm, in the form of (x, y, z) coordinates (Zeng et al., 2022). This dataset was then imported into Originlab (OriginLab, Hampton, MA, United States) for the generation of a heat map of the fracture line.

The integration of fracture heat mapping into the E−3D software (Central South University, Changsha, China) has proven beneficial. The procedure for generating a fracture heat map using this software is straightforward and efficient. Initially, the fracture line is delineated in 3-matic and subsequently exported in txt format. This file is then imported into E−3D, where the heat map is produced via the Fracture Line Analysis module. Due to its user-friendly interface, E−3D software has gained popularity for creating fracture heat maps. In addition, Amira (Thermo Fisher Scientific, Waltham, MA, USA) can also generate fracture thermograms, as reported by Mys et al. (2023).

The position of the quadrilateral plate is shown in Figure 8A. Yang et al. (2018) first conducted a fracture mapping study of quadrilateral plate fractures, analyzing 238 cases and producing a 2D fracture map. The quadrilateral plate was divided into two sections, posterior “A” and anterior “B” (Figure 8B), demarcated by a line from the ischial spine to the iliopubic eminence. They discovered that 65% of the fractures intersected both sections. The fractures were further classified into three categories: those crossing the upper border of both zones (115 cases, 48%), those approximately perpendicular to the inner portion of the arch (110 cases, 46%), and those extending from the upper segment to the arch (60 cases, 25%). In 2019, Yang et al. (2019b) further analyzed the morphological characteristics in quadrilateral region for different acetabular fractures. Their research revealed that the fracture lines in double-column fractures predominantly occurred at the upper and posterior regions of the quadrilateral plate. Conversely, the fracture lines associated with transverse and posterior wall fractures were primarily located in the posterior region. T-shaped acetabular fractures (T-SAF) displayed a relatively uniform distribution of fracture lines across the region. In posterior column fractures, the fracture lines were primarily situated in the middle region, whereas in anterior column fractures, they were predominantly found in the upper region. Ye et al. (2022a) further investigated the morphology of double-column acetabular fractures within the quadrilateral region, corroborating the findings of Yang et al. Their study indicated that the majority of fractures in the quadrilateral region of double-column acetabular fractures were simple and typically located on the peripheral edges of the quadrilateral plate (Ye K. et al., 2022). The acetabular notch emerged as the most frequently affected area, followed by the posterior superior region, while the central region was seldom impacted (Ye K. et al., 2022). These studies highlight that different types of acetabular fractures exhibit distinct fracture line patterns, which are crucial for accurate diagnosis and effective treatment planning.

Acetabular double-column fracture is usually caused by high energy injury, which is the second most common type of acetabular fracture, accounting for 20% of all acetabular fractures (Giannoudis et al., 2005; Ferguson et al., 2010). Due to the destruction of the normal anatomical structure of the acetabular roof and the loss of reference for fracture reduction and fixation, it is one of the most complex types of acetabular fractures (Yang et al., 2020). Yang et al. (2019a) revealed that the fracture lines in acetabular double-column fractures exhibit a Y-shaped distribution, consisting of three branches: one branch from a point near the anterior superior spine to the ischial spine, followed by a branch from the iliac crest to the acetabular roof, and a branche traversing the posterior wall. Unfortunately, they did not create a map of the acetabular articular surface. The condition of the acetabular articular surface is crucial for reduction purposes. Ye et al. (2023a) retrospectively analyzed 100 patients with double-column acetabular fractures and created 3D fracture maps of these fractures. They found that the fracture lines in double-column acetabular fractures exhibit a “dumbbell-shaped” distribution on the acetabular articular surface. These fracture lines start from the triangular region in the lower anterior aspect, where the superior pubic ramus and acetabular wall intersect, then extend through the junction between fossa and anterior lunate surface, and continue into the posterior region of the acetabulum (Ye et al., 2023a). Contrary to the findings of Ye et al., Yin et al. discovered that the fracture lines in double-column fractures are concentrated in the anterior inferior region of the acetabular articular surface, rather than the posterior superior region. Additionally, Yin et al. observed that the fracture lines outside the acetabular articular surface primarily extended from the iliac crest to the apex of the acetabulum, which is similar to the findings of Yang et al.

Posterior wall fracture is the most common type of acetabular fracture, accounting for approximately 1/4 to 1/3 of all acetabular fractures (Zhang et al., 2014; Patel and Moed, 2017; Ahmed et al., 2018). When the hip joint flexes >60° and withstands sufficient violence, it is easy to cause posterior wall fractures. Zhao et al. (Cho et al., 2021a) conducted an analysis on the fracture morphology of acetabular posterior wall fractures using quantitative measurements of fracture characteristics. The study employed the midpoint of the transverse acetabular ligament as the +180° reference point, with the 0° reference point established perpendicular to the ligament, and designated the posterior acetabular surface as the positive area (Figure 8C). They defined the fracture span of the posterior acetabular surface as the ratio of the vertical distance between the fracture beak and acetabular rim to the total length from edge to rim. They found that 80.6% of fracture lines were distributed between 18.7°–117°. In addition, 61.7% of the fracture lines had a fracture span between 0.6 and 0.9. Tian et al. (2022) based on the injury mechanism of posterior wall fractures. They divided all patients into two groups: group A (both-column fractures and posterior wall fractures) and group B (including posterior column and posterior wall fractures; T-shaped fractures and posterior wall fractures; and transverse fractures and posterior wall fractures) (Tian et al., 2022). They found significant differences in fracture morphology between the two groups: the spatial displacement (farthest 3D distance of the fracture) in group A was smaller than that in group B, the fracture area of the articular surface in group A was larger than that in group B, and the fracture line in group B was smaller than that in group B. Group A is more upwardly distributed. Ye et al. (2023b) analyzed the morphology of the posterior elements of the acetabular both-column fracture. They divided the posterior edge of the posterior column and the corresponding posterior acetabular surface (RAS) into three equal parts (the cephalic, medial, and caudal), and analyzed the distribution of fracture lines on the RAS (Ye et al., 2023b). They found that the extra-articular fracture lines of posterior wall fractures were widely distributed throughout the posterior RAS and the cephalic ilium. The intra-articular fracture line originated from the superior rim of the acetabulum, passed downward through the junction of the lunate surface and the fossa, and finally ended in the lower posterior part of the joint. The area of the fracture line they reported was more extensive than that reported by Cho et al. for isolated posterior wall fractures. Their study found that more than two-thirds of transverse line fractures were located in the cephalic third of the RAS, providing the possibility of fracture reduction using an anterior approach (Ye et al., 2023b).

Oguzkaya et al. (2023b) created a 2D fracture map encompassing all types of acetabular fractures. They classified these fractures into three categories: anterior, posterior, and complex fractures. The study revealed that fracture lines in anterior fractures were primarily horizontal and diagonal, with a higher concentration in the lower anterior region. In contrast, posterior fractures typically exhibited oblique lines situated between the acetabular dome and the mid-portion of the posterior wall. For complex fractures, the fracture lines were predominantly located just above the acetabular fossa, the acetabular dome, and the posterosuperior portion of the acetabulum (Oguzkaya et al., 2023b).

T-SAFs represent a complex category of acetabular fractures, constituting approximately 6% of all such injuries (Scheinfeld et al., 2015). T-SAF is characterized by two principal fracture lines: a transverse line that traverses the acetabulum, detaching it from the ilium, and a vertical line that intersects the transverse line below, extending through the anterior and posterior columns (Letournel, 1980). The configuration of these fracture lines resembles the letter “T,” which accounts for the designation of this fracture type. Ye et al. (2022b) developed a fracture map based on an analysis of 56 cases of T-SAF. Their findings indicated that the transverse fracture line typically meanders without compromising the acetabular roof. In contrast, the vertical fracture line tends to incline either anteriorly or posteriorly along the rim of the acetabular fossa, leading to the separation of the obturator foramen (Ye P. et al., 2022). Furthermore, their research revealed that the transitional zone of bone thickness is the most prevalent site for fractures in this context.

Pure transverse acetabular fractures are classified as a fundamental type in the Judet-Letournel system, characterized by their straightforward geometric fracture patterns. Li et al. (2022a) developed fracture map and heat map for 49 cases of transverse fractures, revealing that the anterior edges of these fractures tend to cluster narrowly, whereas the posterior lines are more widely dispersed (Li J. et al., 2022). Their study also highlighted that the distribution of fracture lines within the joint predominantly occupies the upper and middle thirds of the articular surface, with recurrent fracture lines affecting the weight-bearing dome primarily located in the posterior roof region.

IFs are are a prevalent type of injury among the elderly, with intertrochanteric fractures accounting for up to 50% of these cases (Michelson et al., 1995; Socci et al., 2017; Shi et al., 2023). Because their injuries can result in high morbidity, disability, and mortality, intertrochanteric fractures are currently still considered an “unresolved fracture” type (Haidukewych, 2009; 2010; Valizadeh et al., 2012). Details of the studies on Malleolus fractures are summarized in Table 2.

Fu et al. (2019) employed a 2D parameterized fracture probability heatmap to visualize and analyze 100 cases of IFs. They founded that the intertrochanteric line and intertrochanteric ridge were dense regions for fracture. Additionally, they observed an increased probability of greater trochanter fractures with advancing age (Fu et al., 2019). Osteoporosis increased the probability of IFs, although the fracture distribution was dispersed. Females exhibited greater bone loss, resulting in a wider distribution of IFs. The study by Li et al. (2019) also reached a similar conclusion. They found that most of the fracture lines of intertrochanteric fractures were located at the intertrochanteric line where the iliofemoral ligament is attached, the intertrochanteric ridge, and the greater trochanter where the gluteus medius muscle is attached. Similarly, the study by Li et al. (Li C. et al., 2020; Ren et al., 2023b) showed that the fracture lines were mainly concentrated in the lesser trochanter, greater trochanter, intertrochanteric line and intertrochanteric ridge area.

FNFs are most common in elderly patients and are often caused by low-energy injuries such as falls (Johnell and Kanis, 2006). Femoral neck fractures in young patients are mostly caused by high-energy violent injuries, accounting for only 3% of patients with FNFs (Robinson et al., 1995). With the advancement of imaging technology and equipment, internal fixation materials and designs, treatment concepts and surgical techniques, the treatment effect of f FNFs has been significantly improved (Miller et al., 2014). However, the incidence of complications of femoral neck fractures, especially nonunion and avascular necrosis of the femoral head, remains high (Ly and Swiontkowski, 2008).

Öğümsöğütlü and Kılınçoğlu (2023) analyzed the morphology of 120 cases of FNFs. They divided the anterior and posterior regions of the femoral neck into three zones each (Figure 8D). They found that fractures most commonly affected Zone 4, while Zone 6 was the least susceptible. Zones 2 and 5 had significantly higher involvement in patients under the age of 65 (Öğümsöğütlü and Kılınçoğlu, 2023).

FHFs are rare hip injuries that predominantly occur in young adults and are typically caused by high-energy trauma (Chiron and Reina, 2022). These injuries can occur in isolation or as part of more complex hip injuries (Bernstein et al., 2022). Wu et al. (2022) reported the morphology characteristics of FHFs in 2022. They found that fractures were concentrated in the anterior inferior part of the femoral head surface, and fractures with anterior dislocation were mainly distributed on the upper lateral surface of the femoral head surface (Wu et al., 2022).

DFFs are complex, accounting for 0.5% of all fractures and 6% of femoral fractures (Martinet et al., 2000; Elsoe et al., 2018). patients showed bimodal distribution, with high-energy trauma in young patients and low-energy trauma in elderly patients with osteoporosis (Elsoe et al., 2018).

Both Li et al. (2022b) and Chen et al. (2023) studied the fracture morphology of DFFs and their subtypes, and they reached similar conclusions. The fracture lines of distal femoral fractures were mainly concentrated in the metaphysis, the intercondylar notch and around the patellofemoral joint (Li R. et al., 2022; Chen et al., 2023). Among each subtype, the intercondylar fracture morphology of AO/OTA type B and C fractures was similar, whereas the supracondylar fracture pattern is different in AO/OTA types A and C. The fracture line of AO/OTA type A fracture is mainly located near the metaphysis, especially on the anterior surface of the supracondylar area (Li R. et al., 2022; Chen et al., 2023). In type B fractures, the fracture line was mainly concentrated around the intercondylar notch. The fracture line of type C fracture is mainly concentrated on the lateral side of the intercondylar notch and the metaphysis, with a “Y” shape distribution (Li R. et al., 2022; Chen et al., 2023).

Hoffa fracture is a coronal fracture posterior to one or both condyles of the distal femur, accounting for 8.7%–13% of distal femur fractures and is classified as AO/OTA 33-B3-2 or 33-B3-3 (Nork et al., 2005; Peez et al., 2024) This fracture is of great interest because it can be easily overlooked on routine imaging studies (Nork et al., 2005). Xie et al. (2017) determined the distribution location and frequency of Hoffa fractures through 2-D and 3-D CT technology in 2017. They found that Hoffa fractures most commonly occur in the middle third of the lateral condyle, extending from anteroinferior to posterosuperior and from anterolateral to posteromedial Ding et al. also observed such phenomena (Xie et al., 2017). Furthermore, Ding et al. (Ding et al., 2022) found that the lateral condyle α angle (the angle between the coronal fracture and the posterior condylar axis in the axial plane) and β angle (the angle between the coronal fracture and the femoral posterior cortex in the sagittal plane) in type 3-C3 fractures were significantly smaller than those in Hoffa fractures (B3), while there was no significant difference in the angle of the medial condyle. Additionally, the coronal fracture and articular comminution zone in type 33-C3 were more anteriorly located (Ding et al., 2022).

Previous studies on malleolus fractures have predominantly focused on a subset of these injuries. Among these, fractures of the posterior malleolus are the most prevalent, accounting for approximately 7%–44% of all ankle injuries. These fractures frequently occur in conjunction with injuries to the medial and lateral malleoli or ligament damage, presenting a more severe clinical scenario than malleolus fractures that do not involve the posterior malleolus (Koval et al., 2005; Warner et al., 2014). Regarding the morphology of posterior malleolus fractures, there is a general consensus in the literature. The fracture line typically extends obliquely across the posterolateral articular surface of the distal tibia (Mitchell et al., 2019; Su et al., 2020; Quan et al., 2021; Yu et al., 2021). Mitchell et al. (Mitchell et al., 2019) examined the morphology of posterior malleolus fractures and determined that these fractures, when associated with spiral fractures of the distal tibia, consistently exhibited a posterolateral oblique orientation. Su et al. (2020) observed that medial malleolus fractures were present in all cases of posterior malleolar fractures. Furthermore, they compared supination-external rotation ankle fractures with pronation-external rotation ankle fractures and found that the latter displayed larger posterior tibial fragments. Additionally, there was a significant difference in the distribution of the lateral malleolar fracture line between the two groups; specifically, the lateral malleolar fracture line in pronation-external rotation ankle fractures was positioned higher (Su et al., 2020). Quan et al. (Quan et al., 2021) found that most posterior malleolus fractures are single-fragment fractures, consisting of two different sizes of fragments. The larger fragment always originates from the groove of the posterior tibialis posterior and flexor digitorum longus, and terminating at the middle part of the fibular notch. The smaller fragment mainly originates from one-half of the f posterior malleolus fracture, involving one-fourth of the fibular notch (Quan et al., 2021). Yu et al. (Yu et al., 2021) reported the distribution of posterior malleolar fracture line of supination-external rotation ankle fracture. They found that the average ratio of posterior fracture fragment area to total articular surface area was 14.96%, and most of the fracture lines were posterolateral oblique.

Triplane fractures refer to distal tibial fractures that involve coronal, sagittal, and axial plane fractures of the epiphysis (Figure 8E). (Schnetzler and Hoernschemeyer, 2007) This fracture is specific to adolescents and is often referred to as transitional injury (Wuerz and Gurd, 2013). It is relatively uncommon, accounting for 5%–10% of ankle injuries in adolescents, with slightly higher incidence in boys than girls (Wuerz and Gurd, 2013). In 2018, Hadad et al. (2018) reported on the morphological characteristics of 33 cases of triplane fractures in children. Their findings revealed that metaphyseal fractures typically presented as medial-lateral fracture lines in the posterior metaphysis. Additionally, the epiphyseal fractures predominantly displayed a dense area in an inverted “three-pointed star” configuration. The study further indicated consistent patterns in the trajectory of the fracture lines among the patients. In all 33 cases, at least one fracture line exited through the anterior metaphysis. Specifically, the fracture lines exited through the medial malleolus in 11 cases, entered the posteromedial metaphysis in 24 cases, and exited through the posterolateral metaphysis in 19 cases (Hadad et al., 2018).

Cao et al. (2022) developed 3D fracture map of lateral malleolus fractures to predict ligamentous injuries in supination external rotational malleolus fractures. They divided 148 patients into two groups: an unstable group comprising 41 cases, and a stable group consisting of 107 cases. Their analysis revealed that fracture lines in both groups descended from a posterior-upper to an anterior-lower direction. However, the fracture lines in the unstable group were significantly higher than those in the stable group. Furthermore, they identified fracture height and fracture inclination as critical predictive factors for syndesmotic instability, indicating that increased fracture height and inclination correlate with a greater risk of syndesmotic instability (Cao et al., 2022).

Wang and Chang (2023) described the morphological characteristics of Danis-Weber type B lateral malleolus fracture. They found that the apex of nearly half of type B fractures were not on the posterolateral surface, which could damage the mechanical application of antiglide plates (Wang and Chang, 2023). The steeper fracture line and longer fracture spike indicate that the posterior-medial distribution of the fracture tip was further posterior, indicating a greater severity of violence and accompanying injuries. Additionally, they observed that these morphological features—steeper fracture lines and elongated fracture apices—correlated with a more posterior positioning of the sharp fracture end on the posterior-medial side, further implying greater violence and more severe accompanying injuries (Wang and Chang, 2023).

Most studies on morphology of TPFs have come to a consistent conclusion: the fracture lines are mainly concentrated around the lateral condyle and tibial spine. Kerschbaum et al. (2021) created 2D maps of 271 patients with TPFs. They found that the fracture lines of TPFs were mainly distributed in the lateral column and central column of the tibial plateau. A 3D fracture map study by Yao et al. (2020) reached a similar conclusion, showing that the fracture lines were mainly concentrated at the insertion of the anterior and posterior cruciate ligament and the medial side of the lateral condyle. In addition, Yao et al. proposed a nine-column classification for TPFs based on fracture heatmaps. They think that this classification could encompass all fracture types of the lateral and medial columns. Another 3D fracture map study by Yao et al. (Yao P. et al., 2022) reached a similar conclusion: the dense fracture line was mainly located at the medial edge of the lateral condyle. In addition, the fracture line forms a high-density accumulation area between the intercondylar eminence and the anterior edge of the fibular head. Furthermore, they discovered that different types of TPFs exhibited distinct distribution characteristics of fracture lines. Kabra et al. (2023) conducted a retrospective analysis of imaging data from 228 cases of proximal tibia fractures and observed that the fracture lines were predominantly situated around the tibial tubercle, in the sagittal plane above the lateral plateau, and in the coronal plane above the medial plateau (Kabra et al., 2023).

In addition, several studies have conducted comparative analyses of the fracture morphology among different types of TPFs. Molenaars et al. (Molenaars et al., 2015) created a 2D fracture map for tibial plateau fractures and analyzed the fracture patterns of lateral (Schatzker types I, II, and III), medial (Schatzker types IV), and bicondylar fractures (Schatzker types V and VI) (Molenaars et al., 2015). They observed that most fracture lines of lateral condyle fractures are located on the lateral edge of the lateral condyle, many fracture lines of medial condyle fractures leave from the posterior aspect of the lateral side of the plateau, and fracture lines of bilateral condyle fractures are mainly located on tibial spines and the outer margin of the lateral condyle (Molenaars et al., 2015). Chen et al. (Chen et al., 2016) conducted a retrospective analysis of CT data from 186 cases of tibial plateau fractures. Differing from previous studies, they delineated the most common depression areas and measured the depth of the depressions. They found that the depression area for Schatzker type III fractures was located in the anterolateral tibial plateau, Schatzker type V fracture was most likely to cause posterolateral depression, while Schatzker type VI fracture occurred in the center of the tibial plateau. They also found that there were significant differences in depression depth among different types of fractures, and the order of average depression depth was Schatzker type III, II, V, VI and IV (Chen et al., 2016).

McGonagle et al. (2019) generated 2D fracture maps for TPFs. However, they simplified the fracture lines as straight lines based on the entry and exit coordinates of the fracture lines. In reality, many fracture lines are curved or have an “S” shape, so their fracture maps may not accurately depict the actual fracture morphology.

Yao et al. (2022b) conducted a study in which they generated fracture maps and heatmaps specifically for hyperextension TPFs. Their findings indicated that fracture lines predominantly occurred at the anterior edge of the tibial plateau, with minimal involvement of the posterior articular surface. The primary characteristics of hyperextension tibial plateau fractures included anterior compression and posterior avulsion injuries.

Molenaars et al. (2019) studied the articular coronal fracture angle of posterior medial tibial plateau fragments and found that the average angle for posterior medial articular fractures was 44°, with an average articular surface area encompassing 34% of the entire tibial plateau. There were no significant differences in the morphology of posterior medial fragments among Schatzker type IV, V, and VI fractures (Molenaars et al., 2019).

PHFs account for 5% of all fractures and are more commonly in elderly women with osteoporosis (Bergdahl et al., 2020). Most research on fracture mapping in proximal humerus fractures (PHFs) has been confined to a particular fracture type. Details of the studies on Malleolus fractures are summarized in Table 5. In 2019, Hasan et al. (Hasan et al., 2017) retrospectively collected 3D CT data from 48 cases of complex PHFs, transcribing these fractures onto a proximal humeral template to generate 2D fracture maps. Their findings indicated a high incidence of intra-articular fractures, with 52% involving the articular surface, predominantly located at the greater tuberosity, especially at the junction of the supraspinatus and infraspinatus tendon insertion sites. In contrast, fractures infrequently affected the bicipital groove. However, Misir et al. (2023) observed that all OTA/AO 11C3 fractures of the PHFs involved the bicipital groove, a discrepancy that may be due to variations in average age, gender, and prevalence of osteoporotic fractures among the studied populations. Hasan et al. (2017) were the first to elucidate the relationship between PHF morphology and local bone markers alongside soft tissue attachments. Similarly, Mys et al. (2023) discovered that fracture lines predominantly occurred in the surgical neck and between tuberosities and tendon insertions, with rare involvement of the bicipital groove and rotator cuff tendon insertions. They also noted that fractures of the articular surface were uncommon, contrasting with findings from Hasan et al. and Misir et al., Ren et al. (Ren et al., 2023a) classified PHFs into six groups and created fracture maps for each type, but did not further analyze the morphology and distribution of the fracture lines.

Furthermore, several scholars have reported on the fracture morphology of specific locations in PHFs. Ju et al. (2023) reported on the fracture morphology of greater tuberosity fractures in three-part and four-part PHFs. They found that the fracture morphology of the large tubercle was similar between the three-part and the four-part PHFs. The fracture line had three branches in the greater tubercle, which were located in the front, middle and back of the great tubercle. The fracture lines within the greater tuberosity exhibit three branches, located anteriorly, medially, and posteriorly, resulting in an overall fan-shaped distribution. Guo et al. (2023) classified the bicipital groove (BG) into three sections: the upper 1/3, middle 1/3, and lower 1/3, and analyzed the fracture morphology of the bicipital groove in proximal humerus fractures. They found that bicipital groove fractures were predominantly located in the upper two-thirds of the BG, particularly in the middle third.

Distal humeral fractures make up approximately 2% of elbow fractures in adults (Bégué, 2014). Wang et al. (Wang et al., 2021) were the pioneers in employing 3D mapping to elucidate the morphology of distal humeral fractures. Their research identified that the predominant fracture zones were located in the radial fossa, coronal fossa, olecranon fossa, and the external segment of the trochlea (Wang et al., 2021). These insights are crucial for guiding the selection of treatment plans and the design of surgical fixations.

Radial head fracture is a common elbow injury, with a common mechanism of injury being a fall with an outstretched arm (Duckworth et al., 2012b). Mellema et al. (2016) used quantitative 3D CT reconstruction to determine the fracture line distribution and location of 66 intra-articular radial head fractures. They found that fracture lines were concentrated in the anterolateral quadrant of the radial head, with less accumulation in the posteromedial quadrant. Not only that, they conducted a qualitative and quantitative analysis of the fracture lines and found that the distribution and location of fracture lines did not differ between specific fracture patterns in the elbow. Details of the studies on Malleolus fractures are summarized in Table7.

DRFs are one of the most common fracture types in the emergency department, with intra-articular fracture accounting for approximately 25% of all DRF (Gordon et al., 2003). Misir et al. (2018) drew 2D fracture maps of 34 cases of OTA/AO 23C3 fractures. They found that the fracture lines were mainly distributed in the central area of the distal radius articular surface. However, Li et al. (Li S. et al., 2020) obtained different results from Misir et al. They also created fracture maps for AO/OTA type 23C3 distal radius fractures and found that the fracture lines formed a cross pattern on the articular surface, mainly concentrated in the posterior two-thirds of the joint surface and the dorsal ridge of the lunate fossa. Additionally, Li et al. (Li C. et al., 2020) think that the Melone classification system did not adequately represent all C3 fractures. This finding also reveals the shortcomings of the Melone typing system.

Zhang et al. (2019) created 2D fracture maps and heat maps of intra-articular fractures of the DRFs using simplified 3D CT reconstructions. They found that the highest intensity of fracture lines was distributed in the posterior aspect of the joint, forming an inverted “T” shape region. This finding is consistent with the results reported by Li et al., (Zhang et al., 2019), The keystone projected area, the radial styloid process and the metacarpal radial side articular surface were the least affected by fractures.

In addition, Clarnette et al. (2022) generated 2D fracture maps for 23 cases of volar Lunate facet fractures of the distal radius. They found that the fragments of volar lunate facet fractures were displaced towards the palmar and proximal directions, and the lunate bone remained connected to the displaced fragments (Clarnette et al., 2022). However, this study included only 23 patients, and the reliability of the results is uncertain.