Applying deep learning to quantify empty lacunae in histologic sections of osteonecrosis of the femoral head

Can Ordinary AI-Powered Tools Replace a Clinician-Led Fracture Clinic Appointment?

Introduction Artificial intelligence (AI)-powered tools are increasingly integrated into healthcare. The purpose of the present study was to compare fracture management plans generated by clinicians to those obtained from ChatGPT (OpenAI, San Francisco, CA) and Google Gemini (Google, Inc., Mountain View, CA). Methodology A retrospective comparative analysis was conducted. The study included 70 cases of isolated injuries treated at the authors' institution fracture clinic. Complex, open fractures and non-specific diagnoses were excluded. All relevant clinical details were introduced into ChatGPT and Google Gemini. The AI-generated management plans were compared with actual documented plans obtained from the clinical records. The study focused on treatment recommendations and follow-up strategies. Results In terms of agreement with actual treatment plans, Google Gemini matched in only 13 cases (19%), with disagreements in the remainder of cases due to overgeneralisation, inadequate treatment, and ambiguity. In contrast, ChatGPT matched actual plans in 24 cases (34%), with overgeneralisation being the principal cause for disagreement. The differences between AI-powered tools and actual clinician-led plans were statistically significant (p < 0.001). Conclusion Both AI-powered tools demonstrated significant disagreement with actual clinical management plans. While ChatGPT showed closer alignment to human expertise, particularly in treatment recommendations, both AI engines still lacked the clinical precision required for accurate fracture management. These findings highlight the current limitations of ordinary AI-powered tools and negate their ability to replace a clinician-led fracture clinic appointment.

A Multimodal Transfer Learning Approach for Histopathology and SR-microCT Low-Data Regimes Image Segmentation

Osteocyte-lacunar bone structures are a discerning marker for bone pathophysiology, given their geometric alterations observed during aging and diseases. Deep Learning (DL) image analysis has showcased the potential to comprehend bone health associated with their mechanisms. However, DL examination requires labeled and multimodal datasets, which is arduous with high-dimensional images. Within this context, we propose a method for segmenting osteocytes and lacunae in human bone histopathology and Synchrotron Radiation micro-Computed Tomography (SR-microCT) images, employing a deep U-Net in an intra-domain and multimodal transfer learning setting with a limited number of training images. Our strategy allows achieving 63.92±4.69 and 63.94±4.05 Dice Similarity Coefficient (DSC) osteocytes and lacunae segmentation, while up to 20.38 and 5.86 average DSC improvements over selected baselines even if 44× smaller datasets are employed for training.Clinical relevance-The proposed method analyzes bone histopathologies and SR-microCT images in a multimodal and low-data setting, easing the bone microscale investigations while supporting the study of osteocyte-lacunar pathophysiology.

A Deep-Learning Model for Predicting the Efficacy of Non-vascularized Fibular Grafting Using Digital Radiography

RATIONALE AND OBJECTIVES

To develop a fully automated deep-learning (DL) model using digital radiography (DR) with relatively high accuracy for predicting the efficacy of non-vascularized fibular grafting (NVFG) and identifying suitable patients for this procedure.

MATERIALS AND METHODS

A retrospective analysis was conducted on osteonecrosis of femoral head patients who underwent NVFG between June 2009 and June 2021. All patients underwent standard preoperative anteroposterior (AP) and frog-lateral (FL) DR. Subsequently, the radiographs were pre-processed and labeled based on the follow-up results. The dataset was randomly divided into training and testing datasets. The DL-based prediction model was developed in the training dataset and its diagnostic performance was evaluated using the testing dataset.

RESULTS

A total of 339 patients with 432 hips were included in this study, with a hip preservation success rate of 71.52% as of June 2023. The hips were randomly divided into a training dataset (n = 324) and a testing dataset (n = 108). The ensemble model in predicting the efficacy of NVFG, reaching an accuracy of 78.9%, a precision of 78.7%, a recall of 96.0%, a F1-score of 86.5%, and an area under the curve (AUC) of 0.780. FL views (AUC, 0.71) exhibited better performance compared to AP views (AUC, 0.66).

CONCLUSION

The proposed DL model using DR enables automatic and efficient prediction of NVFG efficacy without additional clinical and financial burden. It can be seamlessly integrated into various clinical scenarios, serving as a practical tool to identify suitable patients for NVFG.

Dragon blood resin ameliorates steroid-induced osteonecrosis of femoral head through osteoclastic pathways

Objective: Dragon's Blood resin (DBR) is a traditional medicinal substance renowned for its diverse pharmacological effects, which consists of potent anti-inflammatory, antioxidant and angiogenic properties. This study aimed to elucidate its therapeutic mechanism in alleviating steroid-induced osteonecrosis of the femoral head (SIONFH). Methods: Techniques such as SPR and LC-MS were employed to identify and analyze the target proteins of DBR in bone marrow macrophages (BMMs). In vitro, BMMs were treated with RANKL and DBR, and TRAcP staining and actin belt staining were utilized to assess osteoclast activity. The inhibitory effects and underlying mechanisms of DBR on osteoclastogenesis and reactive oxygen species (ROS) generation were determined using real-time PCR, western blotting and immunofluorescence staining. An in vivo SIONFH rat model was set up to assess the curative impacts of DBR using micro-CT scanning and pathological staining. Results: Bioinformatic tools revealed a pivotal role of osteoclast differentiation in SIONFH. Proteomic analysis identified 164 proteins binding in BMMs. In vitro assessments demonstrated that DBR hindered osteoclastogenesis by modulating the expression of specific genes and proteins, along with antioxidant proteins including TRX1 and Glutathione Reductase. Notably, the resin effectively inhibited the expression of crucial proteins, such as the phosphorylation of JNK and the nuclear localization of p65 within the TRAF6/JNK and NFκB signaling pathways. In vivo experiments further confirmed that DBR mitigated the onset of SIONFH in rats by curbing osteoclast and ROS activities. Conclusion: These findings underscore the potential of Dragon's Blood as an effective administration for early-stage SIONFH, shedding light on its therapeutic influence on ROS-mediated osteoclastic signaling pathways.

Successful real-world application of an osteoarthritis classification deep-learning model using 9210 knees-An orthopedic surgeon's view

This study aimed to evaluate the performance of a deep-learning model to evaluate knee osteoarthritis using Kellgren-Lawrence grading in real-life knee radiographs. A deep convolutional neural network model was trained using 8964 knee radiographs from the osteoarthritis initiative (OAI), including 962 testing set images. Another 246 knee radiographs from the Far Eastern Memorial Hospital were used for external validation. The OAI testing set and external validation images were evaluated by experienced specialists, two orthopedic surgeons, and a musculoskeletal radiologist. The accuracy, interobserver agreement, F1 score, precision, recall, specificity, and ability to identify surgical candidates were used to compare the performances of the model and specialists. Attention maps illustrated the interpretability of the model classification. The model had a 78% accuracy and consistent interobserver agreement for the OAI (model-surgeon 1 К = 0.80, model-surgeon 2 К = 0.84, model-radiologist К = 0.86) and external validation (model-surgeon 1 К = 0.81, model-surgeon 2 К = 0.82, model-radiologist К = 0.83) images. A lower interobserver agreement was found in the images misclassified by the model (model-surgeon 1 К = 0.57, model-surgeon 2 К = 0.47, model-radiologist К = 0.65). The model performed better than specialists in identifying surgical candidates (Kellgren-Lawrence Stages 3 and 4) with an F1 score of 0.923. Our model not only had comparable results with specialists with respect to the ability to identify surgical candidates but also performed consistently with open database and real-life radiographs. We believe the controversy of the misclassified knee osteoarthritis images was based on a significantly lower interobserver agreement.