Machine-Learning Approach to Differentiation of Benign and Malignant Peripheral Nerve Sheath Tumors: A Multicenter Study

Abstract

BACKGROUND Clinicoradiologic differentiation between benign and malignant peripheral nerve sheath tumors (PNSTs) has important management implications. OBJECTIVE To develop and evaluate machine-learning approaches to differentiate benign from malignant PNSTs. METHODS We identified PNSTs treated at 3 institutions and extracted high-dimensional radiomics features from gadolinium-enhanced, T1-weighted magnetic resonance imaging (MRI) sequences. Training and test sets were selected randomly in a 70:30 ratio. A total of 900 image features were automatically extracted using the PyRadiomics package from Quantitative Imaging Feature Pipeline. Clinical data including age, sex, neurogenetic syndrome presence, spontaneous pain, and motor deficit were also incorporated. Features were selected using sparse regression analysis and retained features were further refined by gradient boost modeling to optimize the area under the curve (AUC) for diagnosis. We evaluated the performance of radiomics-based classifiers with and without clinical features and compared performance against human readers. RESULTS A total of 95 malignant and 171 benign PNSTs were included. The final classifier model included 21 imaging and clinical features. Sensitivity, specificity, and AUC of 0.676, 0.882, and 0.845, respectively, were achieved on the test set. Using imaging and clinical features, human experts collectively achieved sensitivity, specificity, and AUC of 0.786, 0.431, and 0.624, respectively. The AUC of the classifier was statistically better than expert humans (P = .002). Expert humans were not statistically better than the no-information rate, whereas the classifier was (P = .001). CONCLUSION Radiomics-based machine learning using routine MRI sequences and clinical features can aid in evaluation of PNSTs. Further improvement may be achieved by incorporating additional imaging sequences and clinical variables into future models.

Commentary: Machine-Learning Approach to Differentiation of Benign and Malignant Peripheral Nerve Sheath Tumors: A Multicenter Study

“…It complements clinical, histological, and genetic analysis, especially in patients with extensive tumor burden. The authors 20 of this manuscript suggest a strong role for radiomics using standard MRI sequences in the detection of malignant change in PNST. This is still at a nascent stage with limited penetrance of this technology, but it could be immensely valuable, particularly in smaller or remote centers with a scant exposure to such rare tumors; radiomic analysis suggesting malignant change would prompt immediate triage to tertiary centers with specialized capabilities.…”

Malignant and Benign Peripheral Nerve Sheath Tumors in a Single Center

“…Early identification of MPNST can improve screening strategies and guide clinical decision-making. [6][7][8] The gold standard for the diagnosis of MPNSTs should be based on the pathological findings; however, routine biopsy is impractical because MPNSTs are rare, and biopsies may be painful and cause persistent nerve damage. 1,9 Ultrasonography (US) and magnetic resonance imaging (MRI) are the best imaging techniques for describing the characteristics of soft tissue tumors and guiding the clinical diagnosis and treatment of patients.…”

“…Early detection of MPNSTs is crucial for their treatment management, which involves complete resection with negative margins and active multidisciplinary treatment, and differs significantly from the approach for nonsurgical benign peripheral nerve sheath tumors (BPNST), which do not require surgery. Early identification of MPNST can improve screening strategies and guide clinical decision‐making 6–8 …”

Commentary: Natural History of Brachial Plexus, Peripheral Nerve, and Spinal Schwannomas

“…In fact, recent studies have shown that machine learning models can surpass clinical experts in identifying benign vs malignant peripheral nerve sheath tumors. 9,10…”

“…In fact, recent studies have shown that machine learning models can surpass clinical experts in identifying benign vs malignant peripheral nerve sheath tumors. 9,10 Overall, the published work is a valuable foundational study that appropriately characterizes the natural history of extracranial schwannomas in a large cohort. Focused efforts using precision medicine techniques will be needed to translate the population-level estimates derived by the authors to the care of individual patients.…”