Comparison of five convolutional neural networks for predicting osteoporosis based on mandibular cortical index on panoramic radiographs

OBJECTIVES

The aim of the present study was to compare five convolutional neural networks for predicting osteoporosis based on mandibular cortical index (MCI) on panoramic radiographs.

METHODS

Panoramic radiographs of 744 female patients over 50 years of age were labeled as C1, C2, and C3 depending on the MCI. The data of the present study were reviewed in different categories including (C1, C2, C3), (C1, C2), (C1, C3), and (C1, (C2 +C3)) as two-class and three-class predictions. The data were separated randomly as 20% test data, and the remaining data were used for training and validation with fivefold cross-validation. AlexNET, GoogleNET, ResNET-50, SqueezeNET, and ShuffleNET deep-learning models were trained through the transfer learning method. The results were evaluated by performance criteria including accuracy, sensitivity, specificity, F1-score, AUC, and training duration. The Gradient-Weighted Class Activation Mapping (Grad-CAM) method was applied for visual interpretation of where deep-learning algorithms gather the feature from image regions.

RESULTS

The dataset (C1, C2, C3) has an accuracy rate of 81.14% with AlexNET; the dataset (C1, C2) has an accuracy rate of 88.94% with GoogleNET; the dataset (C1, C3) has an accuracy rate of 98.56% with AlexNET; and the dataset (C1,(C2+C3)) has an accuracy rate of 92.79% with GoogleNET.

CONCLUSION

The highest accuracy was obtained in the differentiation of C3 and C1 where osseous structure characteristics change significantly. Since the C2 score represent the intermediate stage (osteopenia), structural characteristics of the bone present behaviors closer to C1 and C3 scores. Therefore, the data set including the C2 score provided relatively lower accuracy results.

Osteoporosis detection in panoramic radiographs using a deep convolutional neural network-based computer-assisted diagnosis system: a preliminary study

OBJECTIVES

To evaluate the diagnostic performance of a deep convolutional neural network (DCNN)-based computer-assisted diagnosis (CAD) system in the detection of osteoporosis on panoramic radiographs, through a comparison with diagnoses made by oral and maxillofacial radiologists.

METHODS

Oral and maxillofacial radiologists with >10 years of experience reviewed the panoramic radiographs of 1268 females {mean [± standard deviation (SD)] age: 52.5 ± 22.3 years} and made a diagnosis of osteoporosis when cortical erosion of the mandibular inferior cortex was observed. Among the females, 635 had no osteoporosis [mean (± SD) age: 32.8 ± SD 12.1 years] and 633 had osteoporosis (72.2 ± 8.5 years). All panoramic radiographs were analysed using three CAD systems, single-column DCNN (SC-DCNN), single-column with data augmentation DCNN (SC-DCNN Augment) and multicolumn DCNN (MC-DCNN). Among the radiographs, 200 panoramic radiographs [mean (± SD) patient age: 63.9 ± 10.7 years] were used for testing the performance of the DCNN in detecting osteoporosis in this study. The diagnostic performance of the DCNN-based CAD system was assessed by receiver operating characteristic (ROC) analysis.

RESULTS

The area under the curve (AUC) values obtained using SC-DCNN, SC-DCNN (Augment) and MC-DCNN were 0.9763, 0.9991 and 0.9987, respectively.

CONCLUSIONS

The DCNN-based CAD system showed high agreement with experienced oral and maxillofacial radiologists in detecting osteoporosis. A DCNN-based CAD system could provide information to dentists for the early detection of osteoporosis, and asymptomatic patients with osteoporosis can then be referred to the appropriate medical professionals.

Automatic detection of osteoporosis based on hybrid genetic swarm fuzzy classifier approaches

OBJECTIVES

This study proposed a new automated screening system based on a hybrid genetic swarm fuzzy (GSF) classifier using digital dental panoramic radiographs to diagnose females with a low bone mineral density (BMD) or osteoporosis.

METHODS

The geometrical attributes of both the mandibular cortical bone and trabecular bone were acquired using previously developed software. Designing an automated system for osteoporosis screening involved partitioning of the input attributes to generate an initial membership function (MF) and a rule set (RS), classification using a fuzzy inference system and optimization of the generated MF and RS using the genetic swarm algorithm. Fivefold cross-validation (5-FCV) was used to estimate the classification accuracy of the hybrid GSF classifier. The performance of the hybrid GSF classifier has been further compared with that of individual genetic algorithm and particle swarm optimization fuzzy classifiers.

RESULTS

Proposed hybrid GSF classifier in identifying low BMD or osteoporosis at the lumbar spine and femoral neck BMD was evaluated. The sensitivity, specificity and accuracy of the hybrid GSF with optimized MF and RS in identifying females with a low BMD were 95.3%, 94.7% and 96.01%, respectively, at the lumbar spine and 99.1%, 98.4% and 98.9%, respectively, at the femoral neck BMD. The diagnostic performance of the proposed system with femoral neck BMD was 0.986 with a confidence interval of 0.942-0.998. The highest mean accuracy using 5-FCV was 97.9% with femoral neck BMD.

CONCLUSIONS

The combination of high accuracy along with its interpretation ability makes this proposed automatic system using hybrid GSF classifier capable of identifying a large proportion of undetected low BMD or osteoporosis at its early stage.

Osteoporosis Detection Using Machine Learning Techniques and Feature Selection

Osteoporosis is a disease of bones that leads to an increased risk of fracture and it is characterized by low bone mineral density and micro-architectural deterioration of bone tissue. In this article, the dataset consists of 3426 subjects (1083 pathological and 2343 healthy cases) whose diagnosis was based on laboratory and osteal bone densitometry examination. In all cases, four diagnostic factors for osteoporosis risk prediction, namely age, sex, height and weight were stored for later evaluation with the selected classifiers. In order to categorize subjects into two classes (osteoporosis, nonosteoporosis), twenty machine learning techniques were assessed, based on their popularity and frequency in biomedical engineering problems. All classifiers have been evaluated using the wellknown 10-fold cross validation method and the results are reported analytically. In addition, a feature selection method identified that with the use of only two diagnostic factors (age and weight), similar performance could be achieved. The scope of the proposed exhaustive methodology is to assist therapists in osteoporosis prediction, avoiding unnecessary further testing with bone densitometry.

Relationship between findings of mandibular cortical bone in inferior border and bone mineral densities of lumbar vertebrae in postmenopausal women

It is important to assess the general bone condition in dental implant treatment. The relationships between the bone mineral densities (BMDs) of lumbar vertebrae and mandibular cortical bone condition in the inferior border using multi-slice computed tomography (CT) were assessed in postmenopausal women. If a strong correlation between them is obtained, the mandibular cortical bone condition may be useful to evaluate the general bone condition. Twenty-two postmenopausal women were enrolled in this investigation. The maximum CT value and width of the mandibular cortical bone (MCW-MSCT) were measured, and the mandibular cortical bone index (MCI-MSCT) was classified using multi-slice CT. BMDs of lumbar vertebrae were measured using dual energy X-ray absorptiometry (DXA), and then the percentage of the young adult mean (YAM) was analyzed. The correlations were investigated between the mandibular cortical bone condition and values of lumbar vertebrae. Weak correlations were observed between MCW-MSCT and the percentage of YAM. Also, significant differences in the percentage of YAM were noted between types of MCI-MSCT. The mandibular cortical bone index (MCI-MSCT) in the inferior border using multi-slice CT may be applicable to evaluate the influence of the general bone condition.

Triage screening for osteoporosis in dental clinics using panoramic radiographs

Many patients with osteoporosis go undiagnosed because typically no symptoms are present before a fracture. Triage screening to refer patients to appropriate medical professionals for further investigation would be useful to address the increase in the incidence of osteoporotic fractures. Dental clinics may offer a new triage screening pathway because dentists frequently take radiographs of bones in the course of dental treatment. A major premise for such triage screening in dental clinics is that dentists can readily use a screening tool in their dental practice. For example, cortical width and shape of the mandible detected on panoramic radiographs may be appropriate indices for triaging individuals with osteoporosis. To date, several investigators have demonstrated significant associations between cortical indices on panoramic radiographs and bone mineral density of the skeleton generally, such as the spine and femur, biochemical markers of bone turnover and risk of osteoporotic fractures. Further, in two recent Japanese clinical trials, approximately 95% of women who were identified by trained dentists in their clinics using cortical shape findings did have osteopenia or osteoporosis. These findings support the possibility that dental clinics may offer a new triage platform to identify individuals with otherwise undetected osteoporosis.