MATERIALS AND METHODS: Seven hundred fifty-four radiomics-based features were extracted from 1732 scans derived from the TICH-2 multicentre clinical trial. Features were harmonised and a correlation-based feature selection was applied. Different elastic-net parameterisations were tested to assess the predictive performance of the selected radiomics-based features using grid optimisation. For comparison, the same procedure was run using radiological signs and clinical factors separately. Models trained with radiomics-based features combined with radiological signs or clinical factors were tested. Predictive performance was evaluated using the area under the receiver operating characteristic curve (AUC) score.

RESULTS: The optimal radiomics-based model showed an AUC of 0.693 for haematoma expansion and an AUC of 0.783 for poor functional outcome. Models with radiological signs alone yielded substantial reductions in sensitivity. Combining radiomics-based features and radiological signs did not provide any improvement over radiomics-based features alone. Models with clinical factors had similar performance compared to using radiomics-based features, albeit with low sensitivity for haematoma expansion. Performance of radiomics-based features was boosted by incorporating clinical factors, with time from onset to scan and age being the most important contributors for haematoma expansion and poor functional outcome prediction, respectively.

CONCLUSION: Radiomics-based features perform better than radiological signs and similarly to clinical factors on the prediction of haematoma expansion and poor functional outcome. Moreover, combining radiomics-based features with clinical factors improves their performance.

PURPOSE: To evaluate the accuracy, safety, and diagnostic outcome of fluoroscopic guided and CT transpedicular biopsy techniques.

STUDY DESIGN: Prospective randomized trial.

PATIENT SAMPLE: Sixty consecutive patients with clinical symptoms and radiological features suggestive of spinal infection or malignancy were recruited and randomized into fluoroscopic or CT guided spinal biopsy groups. Both groups were similar in terms of patient demographics, distribution of spinal infections and malignancy cases, and the level of biopsies.

OUTCOME MEASURES: The primary outcome measure was diagnostic accuracy of both methods, determined based on true positive, true negative, false positive, and false negative biopsy findings. Secondary outcome measures included radiation exposure to patients and doctors, complications, and postbiopsy pain score.

METHODS: A transpedicular approach was performed with an 8G core biopsy needle. Specimens were sent for histopathological and microbiological examinations. Diagnosis was made based on biopsy results, clinical criteria and monitoring of disease progression during a 6-month follow up duration. Clinical criteria included presence of risk factors, level of inflammatory markers and magnetic resonance imaging findings. Radiation exposure to patients and doctors was measured with dosimeters.

RESULTS: There was no significant difference between the diagnostic accuracy of fluoroscopic and CT guided spinal biopsy (p=0.67) or between the diagnostic accuracy of spinal infection and spinal tumor in both groups (p=0.402 for fluoroscopy group and p=0.223 for CT group). Radiation exposure to patients was approximately 26 times higher in the CT group. Radiation exposure to doctors in the CT group was approximately 2 times higher compared to the fluoroscopic group if a lead shield was not used. Lead shields significantly reduced radiation exposure to doctors anywhere from 2 to 8 times. No complications were observed for either group and the differences in postbiopsy pain scores were not significant.

METHOD: Two large datasets, including 1110 3D CT images, were split into five segments of 20% each. Each dataset's first 20% segment was separated as a holdout test set. 3D-CNN training was performed with the remaining 80% from each dataset. Two small external datasets were also used to independently evaluate the trained models.

RESULTS: The total combination of 80% of each dataset has an accuracy of 91% on Iranmehr and 83% on Moscow holdout test datasets. Results indicated that 80% of the primary datasets are adequate for fully training a model. The additional fine-tuning using 40% of a secondary dataset helps the model generalize to a third, unseen dataset. The highest accuracy achieved through transfer learning was 85% on LDCT dataset and 83% on Iranmehr holdout test sets when retrained on 80% of Iranmehr dataset.