Dual-Stage Deeply Supervised Attention-Based Convolutional Neural Networks for Mandibular Canal Segmentation in CBCT Scans

Abstract

Accurate segmentation of mandibular canals in lower jaws is important in dental implantology. Medical experts manually determine the implant position and dimensions from 3D CT images to avoid damaging the mandibular nerve inside the canal. In this paper, we propose a novel dual-stage deep learning-based scheme for the automatic segmentation of the mandibular canal. In particular, we first enhance the CBCT scans by employing the novel histogram-based dynamic windowing scheme, which improves the visibility of mandibular canals. After enhancement, we designed 3D deeply supervised attention UNet architecture for localizing the Volumes Of Interest (VOIs), which contain the mandibular canals (i.e., left and right canals). Finally, we employed the Multi-Scale input Residual UNet (MSiR-UNet) architecture to segment the mandibular canals using VOIs accurately. The proposed method has been rigorously evaluated on 500 and 15 CBCT scans from our dataset and from the public dataset, respectively. The results demonstrate that our technique improves the existing performance of mandibular canal segmentation to a clinically acceptable range. Moreover, it is robust against the types of CBCT scans in terms of field of view.

SISTR: Sinus and Inferior alveolar nerve Segmentation with Targeted Refinement on Cone Beam Computed Tomography images

“…This was based on 256 synthetic annotations, 68 actual annotations, and an internal test set of 15 CBCTs, although limited by the use of a single data source. In contrast, Usman et al 24 reported Dice scores of 0.751 and 0.77 on an internal test set of 500 CBCTs from a unique center and on Cipriano et al’s dataset 25 , respectively, employing a development set of 500 densely annotated CBCTs from a single center. While these results are noteworthy, they are in a different range of magnitude compared to our achieved Dice score of 0.843 (range: 0.810-0.856), underscoring the potential superiority of our method.…”

“…The literature acknowledges the complexity of IAN segmentation, largely due to considerable inter-observer and intra-observer variability; for instance, Järnstedt et al 14 reported an inter-observer variability of 0.77 mm, benchmarked against a gold standard set by experienced radiologists. Comparative studies like Cipriano et al’s 25 and Usman et al’s 24 used internal test sets with limited diversity, primarily from one or two centers, often overlapping with their development sets. This lack of external validation and diversity in test datasets limits the generalizability of their findings.…”

“…The literature supports a two-stage approach, starting with anatomical region localization, as more suitable given the technical and computational challenges. 27,19,24,30,17,26 . This method avoids the need to downsample the entire volume for GPU compatibility, which would compromise the segmentation detail necessary for precise implant planning.…”

“…In the first stage of our framework, we opt for a regression-based method to detect distinct instances through predictions of offset maps to the nearest anatomical region’s centroids. This strategy allows us to avoid the complexities and limitations of instance segmentation, which requires precise class-specific predictions, and semantic segmentation, which, while effective for distinguishing IAN and sinus regions due to their spatial separation, is inadequate for accurately segmenting closely positioned teeth without extensive post-processing for object instance separation 24 .…”

“…Their multi-tier training process involves initial training on sparse data, followed by pretraining on synthetically generated dense annotations, and final finetuning on actual dense annotations, making integration into AI products challenging. The double stage approach of Usman et al 24 , while effective for IAN and sinus due to their natural separability in the oral cavity, may not extend as well to other anatomical structures such as the teeth where distinct separability is lacking, limiting its broader applicability.…”