Search

The diagnostic rate of 99mTc-MDP (methylene diphosphonate) bone scans for cancer bone metastasis is limited by image resolution and high similarity between images. This study aims to enhance the accuracy and precision of bone metastasis detection by applying fusion CNN (convolutional neural network) methods combined with artificial intelligence (AI) deep learning algorithms. In this retrospective study, a total of 1680 whole-body images were collected, including 1002 normal and 678 metastatic cases. Five popular CNN methods were used for transfer learning, extracting features from fully connected layers to create a new set of fusion features. The hyperparameters included batch size, epoch size, and learning rates (0.001). The dataset was divided into 70%, 20%, and 10% for training, validation, and testing of the CNN models, respectively. Machine learning models, such as SVM (support vector machine), Naive Bayes, and logistic regression, were trained using the fusion features. Classification performance was evaluated using accuracy, precision, and Kappa value. The fusion CNN combined with the Naive Bayes classifier achieved the highest accuracy, precision, and Kappa value of 0.982, 0.990, and 0.963, respectively, in the testing set compared to other methods. This study demonstrated that fusion CNN methods combined with machine learning techniques can achieve outstanding performance in classifying normal and bone metastasis images, enhancing the diagnostic potential of 99mTc-MDP bone scans. [ABSTRACT FROM AUTHOR]

Objective: To measure the level of shared decision-making (SDM) in primary care consultations in Malaysia, a multicultural, middle-income developing country.Methods: A cross-sectional study was conducted in an urban, public primary care clinic. Convenience sampling was used to recruit participants, and audio-recorded consultations were scored for SDM levels by two independent raters using the OPTION tool. Univariate and multivariate analysis was conducted to determine factors significantly associated with SDM levels.Results: 199 patients and 31 doctors participated. Mean consultation time was 14.3 min (+ SD 5.75). Patients' age ranged from 18 to 87 years (median age of 57.5 years). 52.8 % of patients were female, with three main ethnicities (Malay, Chinese, Indian). The mean OPTION score was found to be 7.8 (+ SD 3.31) out of 48. After a multivariate analysis, only patient ethnicity (β= -0.142, p < 0.05) and increased consultation time (β = 0.407, p < 0.01) were associated with higher OPTION scores.Conclusions: Patients in Malaysia experience extremely poor levels of SDM in general practice. Higher scores were associated with increased consultation time and patient ethnicity.Practice Implications: Malaysian general practitioners should aim to develop and practice cultural competency skills to avoid biased SDM practice towards certain ethnicities. [ABSTRACT FROM AUTHOR]