Reproducibility in Radiomics: A Comparison of Feature Extraction Methods and Two Independent Datasets.

Radiomics involves the extraction of information from medical images that are not visible to the human eye. There is evidence that these features can be used for treatment stratification and outcome prediction. However, there is much discussion about the reproducibility of results between different studies. This paper studies the reproducibility of CT texture features used in radiomics, comparing two feature extraction implementations, namely the MATLAB toolkit and Pyradiomics, when applied to independent datasets of CT scans of patients: (i) the open access RIDER dataset containing a set of repeat CT scans taken 15 min apart for 31 patients (RIDER Scan 1 and Scan 2, respectively) treated for lung cancer; and (ii) the open access HN1 dataset containing 137 patients treated for head and neck cancer. Gross tumor volume (GTV), manually outlined by an experienced observer available on both datasets, was used. The 43 common radiomics features available in MATLAB and Pyradiomics were calculated using two intensity-level quantization methods with and without an intensity threshold. Cases were ranked for each feature for all combinations of quantization parameters, and the Spearman's rank coefficient, rs, calculated. Reproducibility was defined when a highly correlated feature in the RIDER dataset also correlated highly in the HN1 dataset, and vice versa. A total of 29 out of the 43 reported stable features were found to be highly reproducible between MATLAB and Pyradiomics implementations, having a consistently high correlation in rank ordering for RIDER Scan 1 and RIDER Scan 2 (rs > 0.8). 18/43 reported features were common in the RIDER and HN1 datasets, suggesting they may be agnostic to disease site. Useful radiomics features should be selected based on reproducibility. This study identified a set of features that meet this requirement and validated the methodology for evaluating reproducibility between datasets.

Sinonasal phosphaturic mesenchymal tumour: radiation oncologist's perspective.

Tumour-induced osteomalacia is a rare cause of osteomalacia, the majority of which is of mesenchymal origin. Oncogenic osteomalacia is a potentially curable condition caused by phosphaturic mesenchymal tumours. We present the case of a woman in her 30s with a sinonasal phosphaturic mesenchymal tumour, treated with surgical excision followed by adjuvant intensity-modulated radiotherapy and subsequent adjuvant chemotherapy. The patient experienced minimal adverse effects during radiation. There was good local control and cosmetic outcomes with no radiation-related toxicity at a follow-up period of 32 months.

Multi-centre radiomics for prediction of recurrence following radical radiotherapy for head and neck cancers: Consequences of feature selection, machine learning classifiers and batch-effect harmonization.

Background and purpose: Radiomics models trained with limited single institution data are often not reproducible and generalisable. We developed radiomics models that predict loco-regional recurrence within two years of radiotherapy with private and public datasets and their combinations, to simulate small and multi-institutional studies and study the responsiveness of the models to feature selection, machine learning algorithms, centre-effect harmonization and increased dataset sizes. Materials and methods: 562 patients histologically confirmed and treated for locally advanced head-and-neck cancer (LA-HNC) from two public and two private datasets; one private dataset exclusively reserved for validation. Clinical contours of primary tumours were not recontoured and were used for Pyradiomics based feature extraction. ComBat harmonization was applied, and LASSO-Logistic Regression (LR) and Support Vector Machine (SVM) models were built. 95% confidence interval (CI) of 1000 bootstrapped area-under-the-Receiver-operating-curves (AUC) provided predictive performance. Responsiveness of the models' performance to the choice of feature selection methods, ComBat harmonization, machine learning classifier, single and pooled data was evaluated. Results: LASSO and SelectKBest selected 14 and 16 features, respectively; three were overlapping. Without ComBat, the LR and SVM models for three institutional data showed AUCs (CI) of 0.513 (0.481-0.559) and 0.632 (0.586-0.665), respectively. Performances following ComBat revealed AUCs of 0.559 (0.536-0.590) and 0.662 (0.606-0.690), respectively. Compared to single cohort AUCs (0.562-0.629), SVM models from pooled data performed significantly better at AUC = 0.680. Conclusions: Multi-institutional retrospective data accentuates the existing variabilities that affect radiomics. Carefully designed prospective, multi-institutional studies and data sharing are necessary for clinically relevant head-and-neck cancer prognostication models.

Feasibility, safety and oncological outcomes of minimally invasive oesophagectomy following neoadjuvant chemoradiotherapy for oesophageal squamous cell carcinoma - Experience from a tertiary care centre.

Background: Neoadjuvant chemoradiotherapy (nCRT) has improved the resectability and survival of operable oesophageal squamous cell carcinoma (OSCC). We aimed to study if nCRT for OSCC makes minimally invasive oesophagectomy (MIO) technically more challenging and if the peri-operative and oncological outcomes are acceptable for MIO following nCRT. Materials and Methods: A retrospective review of patients with OSCC (cT1-2N1-2, cT3-4aN0-2) treated with nCRT and MIO between 2013 and 2019 was performed. The operative details including the technical difficulty in tumour dissection and lymphadenectomy, the post-operative complications and oncological outcomes were studied. Results: Seventy-five patients (male:female - 50:25; mean [range] age - 55.49 ± 8.43 [22-72] years; stage II - 34.7%; stage III - 37.3%; stage IVA - 28.0%) were enrolled. The concurrent chemotherapy course was completed by 25.3% of patients and the most common reason limiting the completion of chemotherapy was neutropaenia (66.0%). A thoraco-laparoscopic (n = 60) or hybrid (n = 15) McKeown's oesophagectomy with a two-field lymphadenectomy was performed. The increased surgical difficulty was reported in 41 (54.7%) patients, particularly for mid-thoracic tumours and tumours exhibiting incomplete response. The 30-day overall and major complication rate was 48.0% and 20.0%, respectively, and there was no mortality. The rate of R0 resection, pathological complete response and median lymph nodal yield were 93.3%, 48% and 8 (range: 1-25), respectively. The mean overall survival (OS) was 62.2 months (95% confidence interval [CI]: 52.6-71.8) and recurrence-free survival (RFS) was 53.5 months (95% CI: 43.5-63.5). The 1-, 2- and 3-year OS and RFS were 89.5%, 78.8% and 64.4% and 71.1%, 61.3% and 56.6%, respectively. Conclusion: Minimally invasive McKeown's oesophagectomy is feasible and safe in patients with OSCC receiving nCRT. The radiation component of nCRT increases the degree of operative difficulty, especially in relation to the supracarinal dissection and lymphadenectomy. However, this drawback did not adversely affect the short-term surgical or the long-term oncological outcomes.

Concurrent chemo-irradiation with weekly cisplatin and paclitaxel in the treatment of locally advanced squamous cell carcinoma of cervix: a phase II study.

BACKGROUND: Cervical cancer is the most common gynecological cancer in Indian women. This study was initiated to assess whether the combination of paclitaxel and cisplatin with radiation was feasible in Indian women. AIMS AND OBJECTIVES: The aim of this study was to assess the immediate tumor response and toxicity of weekly cisplatin and paclitaxel along with radiotherapy in the treatment of cervical cancer. MATERIALS AND METHODS: Women with primary untreated squamous cell carcinoma of the cervix with FIGO stages IB2 to IIIB were treated with weekly injections of cisplatin 30 mg/m2 and paclitaxel 40 mg/m2 for 4 weeks along with radiotherapy. A total of 25 patients were enrolled in this study. Disease was assessed prior to treatment by pelvic examination and contrast enhanced computed tomography scan of the abdomen and pelvis. Response was assessed 6 weeks after completion of treatment using the same parameters. Clinical and radiological response was documented. The toxicity was assessed and was graded using the common toxicity criteria Version 3.0. Intention to treat analysis was used when reporting results. RESULTS: A total of 23 patients completed the intended treatment. There was a complete response rate of 88%, 12% were not available for response assessment. The major toxicity was Grade 3 diarrhea (48%). The mean duration of treatment was 58 days. CONCLUSIONS: Combination chemotherapy with cisplatin and paclitaxel along with radiotherapy in patients with locally advanced squamous cell carcinoma of cervix had a high incidence of acute toxicity. There was no increase in immediate tumor response and progression free survival with this treatment regimen. Hence, this regimen offers no added benefit when compared to the chemo radiation with cisplatin alone.