Fully Automated Volumetric Measurement of Malignant Pleural Mesothelioma from Computed Tomography Images by Deep Learning: Preliminary Results of an Internal Validation

Anderson, Owen; Kidd, Andrew; Goatman, Keith A; Weir, Alexander; Voisey, Jeremy P.; Dilys, Vismantas; Siebert, J. Paul; Blyth, Kevin G.

doi:10.5220/0008976100640073

Cited by 6 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Volumetric measurement has been recognized as a standard reference for tumor assessment (7,8,10,16,23). Some studies have compared volume with RECIST 1.1 measurement and have shown acceptable reliability and validity between them (8,24).…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

CT Evaluation of Lymph Nodes That Merge or Split during the Course of a Clinical Trial: Limitations of RECIST 1.1

Shafiei

Farhadi

Apolo

et al. 2021

Radiology: Imaging Cancer

View full text Add to dashboard Cite

A ssessment of tumor size change in medical imaging is an essential aspect of the clinical evaluation of cancer treatment. Thus, precise and reliable lesion quantification is essential in response evaluation criteria in clinical oncology. The most common criteria used for tumor assessment remains Response Evaluation Criteria in Solid Tumors (RECIST). RECIST is a series of response criteria initially established in 2000 and revised in 2009 (version 1.1). Per RECIST version 1.1, up to two lymph nodes with short-axis diameter equal to or more than 1.5 cm can be selected as target lesions and be evaluated over time (1,2). Despite substantial revisions made in RECIST 1.1, some vital questions and limitations remain unsolved, including interreader and intrareader variations of measurements, improper method of tumor assessment in the setting of image-based focal treatment (chemoembolization, radiofrequency ablation), and lack of tumor size change with new cancer therapies when necrosis occurs (3-5).Adjacent nodes often merge as the disease progresses. Conversely, conglomerated nodes can split into smaller nodes when they shrink in response to treatment. In merged target lymph nodes, RECIST 1.1 calculates the longest short-axis diameter perpendicular to the longest axis of the resulting node (herein known as "short axis"), while in split target lymph nodes, RECIST 1.1 measures the sum of short-axis diameters of all the resulting nodal fragments (2).Recent advancements in imaging modalities have allowed us to measure the volume of the lesions more easily. The literature indicates that volumetric measurements represent a better prediction and outcome than linear one-dimensional or two-dimensional measurements in response evaluation of cancers (6-10). Volumetric measurement is a surrogate for the number of neoplastic cells and provides a more detailed and accurate evaluation of tumor size compared with RECIST 1.1 (8,11,12).In this study, we compared the RECIST 1.1 measurement in split and merged target lymph nodes with volumetric measurement on CT scans from clinical trials. Materials and Methods PatientsIn this institutional review board-approved Health Insurance Portability and Accountability Act-compliant

show abstract

Section: Discussionmentioning

confidence: 99%

“…Volumetric measurement was considered as the reference standard method for tumor response evaluation (7,8,16). Volumes of the nodes before and after merging or splitting were calculated using Vitrea, and relative percent changes of the volumes were recorded.…”

Section: Shafiei Et Almentioning

confidence: 99%

CT Evaluation of Lymph Nodes That Merge or Split during the Course of a Clinical Trial: Limitations of RECIST 1.1

Shafiei

Farhadi

Apolo

et al. 2021

Radiology: Imaging Cancer

View full text Add to dashboard Cite

show abstract

“…To increase robustness and to provide an estimation of confidence, seven CNN models were trained using a sevenfold division of the training data and the volume measurements ensembled. A more detailed technical description of the method has previously been reported by us in conference proceedings 22…”

Section: Methodsmentioning

confidence: 99%

“…A more detailed technical description of the method has previously been reported by us in conference proceedings. 22 …”

Section: Methodsmentioning

confidence: 99%

Fully automated volumetric measurement of malignant pleural mesothelioma by deep learning AI: validation and comparison with modified RECIST response criteria

Kidd

Anderson

Cowell

et al. 2022

Thorax

Self Cite

View full text Add to dashboard Cite

BackgroundIn malignant pleural mesothelioma (MPM), complex tumour morphology results in inconsistent radiological response assessment. Promising volumetric methods require automation to be practical. We developed a fully automated Convolutional Neural Network (CNN) for this purpose, performed blinded validation and compared CNN and human response classification and survival prediction in patients treated with chemotherapy.MethodsIn a multicentre retrospective cohort study; 183 CT datasets were split into training and internal validation (123 datasets (80 fully annotated); 108 patients; 1 centre) and external validation (60 datasets (all fully annotated); 30 patients; 3 centres). Detailed manual annotations were used to train the CNN, which used two-dimensional U-Net architecture. CNN performance was evaluated using correlation, Bland-Altman and Dice agreement. Volumetric response/progression were defined as ≤30%/≥20% change and compared with modified Response Evaluation Criteria In Solid Tumours (mRECIST) by Cohen’s kappa. Survival was assessed using Kaplan-Meier methodology.ResultsHuman and artificial intelligence (AI) volumes were strongly correlated (validation set r=0.851, p<0.0001). Agreement was strong (validation set mean bias +31 cm3 (p=0.182), 95% limits 345 to +407 cm3). Infrequent AI segmentation errors (4/60 validation cases) were associated with fissural tumour, contralateral pleural thickening and adjacent atelectasis. Human and AI volumetric responses agreed in 20/30 (67%) validation cases κ=0.439 (0.178 to 0.700). AI and mRECIST agreed in 16/30 (55%) validation cases κ=0.284 (0.026 to 0.543). Higher baseline tumour volume was associated with shorter survival.ConclusionWe have developed and validated the first fully automated CNN for volumetric MPM segmentation. CNN performance may be further improved by enriching future training sets with morphologically challenging features. Volumetric response thresholds require further calibration in future studies.

show abstract

“…An alternative method of tumor evaluation is volumetric measurement, which can generate more accurate shape calculations in lesions [ 17 ]. Although volumetric segmentation has been considered the “gold standard” of cancer imaging by some authors, it is time- and resource-intensive, which makes it impractical for routine daily usage in clinical practice [ 18 , 19 , 20 , 21 ]. However, this is a more accurate and reliable method in the study of rare cancers such as ACC where the patient population is limited.…”

Section: Introductionmentioning

confidence: 99%

Tumor Doubling Time Using CT Volumetric Segmentation in Metastatic Adrenocortical Carcinoma

et al. 2021

View full text Add to dashboard Cite

Adrenocortical carcinoma (ACC) is a rare malignancy with an overall unfavorable prognosis. Clinicians treating patients with ACC have noted accelerated growth in metastatic liver lesions that requires rapid intervention compared to other metastatic locations. This study measured and compared the growth rates of metastatic ACC lesions in the lungs, liver, and lymph nodes using volumetric segmentation. A total of 12 patients with metastatic ACC (six male; six female) were selected based on their medical history. Computer tomography (CT) exams were retrospectively reviewed and a sampling of ≤5 metastatic lesions per organ were selected for evaluation. Lesions in the liver, lung, and lymph nodes were measured and evaluated by volumetric segmentation. Statistical analyses were performed to compare the volumetric growth rates of the lesions in each organ system. In this cohort, 5/12 had liver lesions, 7/12 had lung lesions, and 5/12 had lymph node lesions. A total of 92 lesions were evaluated and segmented for lesion volumetry. The volume doubling time per organ system was 27 days in the liver, 90 days in the lungs, and 95 days in the lymph nodes. In this series of 12 patients with metastatic ACC, liver lesions showed a faster growth rate than lung or lymph node lesions.

show abstract

Fully Automated Volumetric Measurement of Malignant Pleural Mesothelioma from Computed Tomography Images by Deep Learning: Preliminary Results of an Internal Validation

Cited by 6 publications

References 0 publications

CT Evaluation of Lymph Nodes That Merge or Split during the Course of a Clinical Trial: Limitations of RECIST 1.1

CT Evaluation of Lymph Nodes That Merge or Split during the Course of a Clinical Trial: Limitations of RECIST 1.1

Fully automated volumetric measurement of malignant pleural mesothelioma by deep learning AI: validation and comparison with modified RECIST response criteria

Tumor Doubling Time Using CT Volumetric Segmentation in Metastatic Adrenocortical Carcinoma

Contact Info

Product

Resources

About