1
|
Xie H, Zhang Y, Dong L, Lv H, Li X, Zhao C, Tian Y, Xie L, Wu W, Yang Q, Liu L, Sun D, Qiu L, Shen L, Zhang Y. Deep learning driven diagnosis of malignant soft tissue tumors based on dual-modal ultrasound images and clinical indexes. Front Oncol 2024; 14:1361694. [PMID: 38846984 PMCID: PMC11153704 DOI: 10.3389/fonc.2024.1361694] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/26/2023] [Accepted: 05/08/2024] [Indexed: 06/09/2024] Open
Abstract
Background Soft tissue tumors (STTs) are benign or malignant superficial neoplasms arising from soft tissues throughout the body with versatile pathological types. Although Ultrasonography (US) is one of the most common imaging tools to diagnose malignant STTs, it still has several drawbacks in STT diagnosis that need improving. Objectives The study aims to establish this deep learning (DL) driven Artificial intelligence (AI) system for predicting malignant STTs based on US images and clinical indexes of the patients. Methods We retrospectively enrolled 271 malignant and 462 benign masses to build the AI system using 5-fold validation. A prospective dataset of 44 malignant masses and 101 benign masses was used to validate the accuracy of system. A multi-data fusion convolutional neural network, named ultrasound clinical soft tissue tumor net (UC-STTNet), was developed to combine gray scale and color Doppler US images and clinic features for malignant STTs diagnosis. Six radiologists (R1-R6) with three experience levels were invited for reader study. Results The AI system achieved an area under receiver operating curve (AUC) value of 0.89 in the retrospective dataset. The diagnostic performance of the AI system was higher than that of one of the senior radiologists (AUC of AI vs R2: 0.89 vs. 0.84, p=0.022) and all of the intermediate and junior radiologists (AUC of AI vs R3, R4, R5, R6: 0.89 vs 0.75, 0.81, 0.80, 0.63; p <0.01). The AI system also achieved an AUC of 0.85 in the prospective dataset. With the assistance of the system, the diagnostic performances and inter-observer agreement of the radiologists was improved (AUC of R3, R5, R6: 0.75 to 0.83, 0.80 to 0.85, 0.63 to 0.69; p<0.01). Conclusion The AI system could be a useful tool in diagnosing malignant STTs, and could also help radiologists improve diagnostic performance.
Collapse
Affiliation(s)
- Haiqin Xie
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Yudi Zhang
- College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, Guangdong, China
| | - Licong Dong
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Heng Lv
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Xuechen Li
- National Engineering Laboratory for Big Data System Computing Technology, Shenzhen University, Shenzhen, China
| | - Chenyang Zhao
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Yun Tian
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Lu Xie
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Wangjie Wu
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Qi Yang
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Li Liu
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Desheng Sun
- Shenzhen Hospital, Peking University, Shenzhen, China
| | - Li Qiu
- West China Hospital, Sichuan University, Chengdu, Sichuan, China
| | - Linlin Shen
- College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, Guangdong, China
| | - Yusen Zhang
- Shenzhen Hospital, Peking University, Shenzhen, China
| |
Collapse
|
2
|
Zhang Y, Zhao C, Lv H, Dong L, Xie L, Tian Y, Wu W, Luo H, Yang Q, Liu L, Sun D, Xie H. Benefit of Using Both Ultrasound Imaging and Clinical Information for Predicting Malignant Soft Tissue Tumors. ULTRASOUND IN MEDICINE & BIOLOGY 2023; 49:2459-2468. [PMID: 37704557 DOI: 10.1016/j.ultrasmedbio.2023.08.005] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 05/16/2023] [Revised: 07/16/2023] [Accepted: 08/08/2023] [Indexed: 09/15/2023]
Abstract
OBJECTIVE Ultrasonography (US) is the primary imaging method for soft tissue tumors (STTs), the diagnostic performance of which still requires improvement. To achieve an accurate evaluation of STTs, we built the diagnostic nomogram for STTs using the clinical and US features of patients with STTs. METHODS A total of 613 patients with 195 malignant and 418 benign STTs were retrospectively recruited. We used a blend of clinical and ultrasonic features, as well as exclusively US features, to develop two distinct diagnostic models for STTs: the clinical-US model and the US-only model, respectively. The two models were evaluated and compared by measuring their areas under the receiver operating characteristic curve (AUC), calibration, integrated discrimination improvement (IDI) and decision curve analysis. The performance of the clinical-US model was also compared with that of two radiologists. RESULTS The clinical-US model had better diagnostic performance than the model based on US imaging features alone (AUCs of the clinical-US and US-only models: 0.95 [0.93-0.97] vs. 0.89 [0.87-0.92], p < 0.001; IDI of the two models: 0.15 ± 0.03, p < 0.001). The clinical-US model was also superior to the two radiologists in diagnosing STTs (AUCs of clinical-US model and two radiologists: 0.95 [0.93-0.97] vs. 0.79 [0.75-0.82] and 0.83 [0.80-0.85], p < 0.001). CONCLUSION The diagnostic model based on clinical and US imaging features had high diagnostic performance in STTs, which could help identify malignant STTs for radiologists.
Collapse
Affiliation(s)
- Yusen Zhang
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Chenyang Zhao
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Heng Lv
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Licong Dong
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Lu Xie
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Yun Tian
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Wangjie Wu
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Haiyu Luo
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Qi Yang
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Li Liu
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Desheng Sun
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China
| | - Haiqin Xie
- Department of Ultrasonography, Peking University Shenzhen Hospital, Shenzhen, China.
| |
Collapse
|
3
|
Guedes A, Oliveira MBDR, Melo ASD, Carmo CCMD. Update in Imaging Evaluation of Bone and Soft Tissue Sarcomas. Rev Bras Ortop 2021; 58:179-190. [DOI: 10.1055/s-0041-1736569] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/16/2020] [Accepted: 07/08/2021] [Indexed: 10/19/2022] Open
Abstract
ResumoA evolução na avaliação por imagens dos sarcomas musculoesqueléticos contribuiu para melhora significativa no prognóstico e na sobrevida dos portadores destas neoplasias. A caracterização precisa destas lesões, mediante utilização das modalidades de imagem mais adequadas a cada condição clínica apresentada, é de suma importância no delineamento da abordagem terapêutica a ser instituída, com impacto direto sobre os desfechos clínicos. O presente artigo busca atualizar o leitor a propósito das metodologias de imagem no contexto da avaliação local e sistêmica dos sarcomas ósseos e das partes moles.
Collapse
Affiliation(s)
- Alex Guedes
- Grupo de Oncologia Ortopédica, Hospital Santa Izabel, Santa Casa de Misericórdia da Bahia, Salvador, BA, Brasil
| | - Marcelo Bragança dos Reis Oliveira
- Serviço de Traumato-ortopedia, Hospital Universitário Clementino Fraga Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brasil
| | - Adelina Sanches de Melo
- Serviço de Medicina Nuclear, Hospital Santa Izabel, Santa Casa da Misericórdia da Bahia, Salvador, BA, Brasil
| | | |
Collapse
|
4
|
Tsukamoto S, Mavrogenis AF, Tanaka Y, Errani C. Imaging of Soft Tissue Tumors. Curr Med Imaging 2021; 17:197-216. [PMID: 32660406 DOI: 10.2174/1573405616666200713183400] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/15/2020] [Revised: 02/08/2020] [Accepted: 06/20/2020] [Indexed: 02/07/2023]
Abstract
Differentiation of malignant from benign soft tissue tumors is challenging with imaging alone, including that by magnetic resonance imaging and computed tomography. However, the accuracy of this differentiation has increased owing to the development of novel imaging technology. Detailed patient history and physical examination remain essential for differentiation between benign and malignant soft tissue tumors. Moreover, measurement only of tumor size based on Response Evaluation Criteria In Solid Tumors criteria is insufficient for the evaluation of response to chemotherapy or radiotherapy. Change in metabolic activity measured by 18F-fluorodeoxyglucose positron emission tomography or dynamic contrast enhanced-derived quantitative endpoints can more accurately evaluate treatment response compared to change in tumor size. Magnetic resonance imaging can accurately evaluate essential factors in surgical planning such as vascular or bone invasion and "tail sign". Thus, imaging plays a critical role in the diagnosis and treatment of soft tissue tumors.
Collapse
Affiliation(s)
- Shinji Tsukamoto
- Department of Orthopaedic Surgery, Nara Medical University, Nara, Japan
| | - Andreas F Mavrogenis
- First Department of Orthopaedics, National and Kapodistrian University of Athens, School of Medicine, Athens, Greece
| | - Yasuhito Tanaka
- Department of Orthopaedic Surgery, Nara Medical University, Nara, Japan
| | - Costantino Errani
- Department Orthopaedic Oncology, IRCCS Istituto Ortopedico Rizzoli, Bologna, Italy
| |
Collapse
|
5
|
Leonard L, Meyer HJ, Surov A. [Imaging characteristics of malignant and benign lesions of skeletal muscle]. Radiologe 2017; 57:1059-1070. [PMID: 29181716 DOI: 10.1007/s00117-017-0323-y] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.1] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/23/2022]
Abstract
There are many different tumors and tumor-like lesions with variable biological behavior that may affect the skeletal musculature. The aim of this study was to review the different intramuscular lesions and to provide a classification based on their radiological patterns. Intramuscular lesions can present as solid, liquid, semiliquid or fat equivalent manifestations and also as diffuse muscle enlargement and muscle calcification. Additionally, lesions with mixed patterns of the aforementioned alterations can also occur. Benign and malignant muscle lesions can often manifest with identical radiological patterns, which is why a certain differentiation is often difficult. A systematic radiological description and when possible assignment with respect to etiology and dignity depending on the patient history is necessary in order to recommend a subsequent histological confirmation or to avoid unnecessary confirmation.
Collapse
Affiliation(s)
- L Leonard
- Klinik für Diagnostische und Interventionelle Radiologie, Universitätsklinikum Leipzig, Liebigstraße 20, 04103, Leipzig, Deutschland
| | - H J Meyer
- Klinik für Diagnostische und Interventionelle Radiologie, Universitätsklinikum Leipzig, Liebigstraße 20, 04103, Leipzig, Deutschland
| | - A Surov
- Klinik für Diagnostische und Interventionelle Radiologie, Universitätsklinikum Leipzig, Liebigstraße 20, 04103, Leipzig, Deutschland.
| |
Collapse
|
6
|
Accuracy and role of contrast-enhanced CT in diagnosis and surgical planning in 88 soft tissue tumours of extremities. Eur Radiol 2015; 26:2400-8. [PMID: 26449561 DOI: 10.1007/s00330-015-4047-y] [Citation(s) in RCA: 6] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/24/2015] [Revised: 09/08/2015] [Accepted: 09/23/2015] [Indexed: 12/25/2022]
Abstract
OBJECTIVES Soft tissue tumours (STT) require accurate diagnosis in order to identify potential malignancies. Preoperative planning is fundamental to avoid inadequate treatments. The role of contrast-enhanced computed tomography (CT) for local staging remains incompletely assessed. Aims of the study were to evaluate CT accuracy in discriminating active from aggressive tumours compared to histology and evaluate the role of CT angiography (CTA) in surgical planning. MATERIALS AND METHODS This retrospective cohort series of 88 cases from 1200 patients (7 %) was locally studied by contrast-enhanced CT and CTA in a referral centre: 74 malignant tumours, 14 benign lesions. Contrast-enhancement patterns and relationship of the mass with major vessels and bone were compared with histology on surgically excised samples. Sensitivity, specificity, positive and negative predictive values (PPV, NPV) were evaluated in discriminating active from aggressive tumours. RESULTS Sensitivity in differentiating aggressive tumours from active lesions was 89 %, specificity 84 %, PPV 90 %, NPV 82 %. The relationship between mass and major vessels/bone was fundamental for surgical strategy respectively in 40 % and in 58 % of malignant tumours. CONCLUSION Contrast-enhanced CT and CTA are effective in differentiating aggressive masses from active lesions in soft tissue and in depicting the relationship between tumour and adjacent bones and major vessels. KEY POINTS • Accurate delineation of vascular and bony involvement preoperatively is fundamental for a correct resection. • CT plays a critical role in differential diagnosis of soft tissue masses. • Contrast-enhanced CT and CT angiography are helpful in depicting tumoral vascular involvement. • CT is optimal for characterization of bone involvement in soft tissue malignancies.
Collapse
|
7
|
Abstract
There are several tumors and tumorlike conditions with variable biological behavior that may involve the skeletal musculature. The aim of this work was to review different intramuscular lesions and to provide a classification of muscle lesions based on their radiological patterns as well as to provide as a pictorial essay the imaging characteristics of typical muscle lesions. Radiologically, intramuscular lesions can manifest as solid masses, liquid or semiliquid masses, fat-containing lesions, diffuse muscle enlargement, and muscle calcifications. Additionally, lesions with mixed patterns can also occur. It is noteworthy that different malignant or benign muscle lesions can manifest with identical radiological patterns.
Collapse
Affiliation(s)
- Alexey Surov
- Department of Radiology, Martin-Luther-University Halle-Wittenberg, Halle, Germany.
| | - Marc-André Weber
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
| |
Collapse
|
8
|
MRI, MDCT features, and clinical outcome of extremity leiomyosarcomas: experience in 47 patients. Skeletal Radiol 2014; 43:615-22. [PMID: 24496586 DOI: 10.1007/s00256-014-1823-8] [Citation(s) in RCA: 13] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 11/14/2013] [Revised: 12/26/2013] [Accepted: 01/08/2014] [Indexed: 02/07/2023]
Abstract
OBJECTIVE To describe MRI, MDCT features, and clinical outcome of extremity leiomyosarcomas (LMS). MATERIALS AND METHODS In this IRB-approved, HIPAA-compliant retrospective study, we included 47 patients (23 women, 24 men; mean age: 55.3 years, range: 17-85 years) with pathologically confirmed extremity LMS seen at our adult tertiary cancer center between 2000 and 2012. MRI/MDCT of primary tumors in 23 patients and follow-up in all patients were reviewed by two radiologists in consensus. Clinical data were extracted from electronic medical records. RESULTS Primary tumors were distributed in bones (6 out of 47), deep soft tissues (24 out of 47), and superficial soft tissues (17 out of 47). On imaging (bone = 4, deep soft tissue = 11, superficial soft tissue = 8), compared with skeletal muscle, they were T1 iso-hypointense and T2 hyperintense. Bone LMS were metaphyseal tumors with cortical destruction (3 out of 4). Deep soft-tissue LMS were large with hemorrhage (7 out of 11) and necrosis (10 out of 11). Superficial soft-tissue LMS were relatively smaller, homogeneously enhancing (6 out of 8) tumors. Distant metastases developed in 32 out of 47 patients (bone LMS [6 out of 6], deep soft-tissue LMS [18 out of 24], superficial soft-tissue LMS [8 out of 17]), commonly to lung (29 out of 47) and bone (14 out of 47). At the time of writing, 22 out of 36 patients (bone LMS [4 out of 6], deep soft-tissue LMS [15 out of 24], superficial soft-tissue LMS [4 out of 17]) have died. There was no statistically significant correlation between metastatic disease and tumor size or grade. CONCLUSION Extremity LMS arise in bones and in the deep and superficial soft tissues, frequently metastasize to the lungs, and have a poor prognosis. Superficial LMS tend to have a better prognosis than bone or deep soft-tissue LMS.
Collapse
|
9
|
MRI-Based Assessment of Safe Margins in Tumor Surgery. Sarcoma 2014; 2014:686790. [PMID: 24701131 PMCID: PMC3950827 DOI: 10.1155/2014/686790] [Citation(s) in RCA: 11] [Impact Index Per Article: 1.1] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/17/2013] [Accepted: 01/19/2014] [Indexed: 11/17/2022] Open
Abstract
Introduction. In surgical oncology, histological analysis of excised tumor specimen is the conventional method to assess the safety of the resection margins. We tested the feasibility of using MRI to assess the resection margins of freshly explanted tumor specimens in rats. Materials and Methods. Fourteen specimen of sarcoma were resected in rats and analysed both with MRI and histologically. Slicing of the specimen was identical for the two methods and corresponding slices were paired. 498 margins were measured in length and classified using the UICC classification (R0, R1, and R2). Results. The mean difference between the 498 margins measured both with histology and MRI was 0.3 mm (SD 1.0 mm). The agreement interval of the two measurement methods was [−1.7 mm; 2.2 mm]. In terms of the UICC classification, a strict correlation was observed between MRI- and histology-based classifications (κ = 0.84, P < 0.05). Discussion. This experimental study showed the feasibility to use MRI images of excised tumor specimen to assess the resection margins with the same degree of accuracy as the conventional histopathological analysis. When completed, MRI acquisition of resected tumors may alert the surgeon in case of inadequate margin and help advantageously the histopathological analysis.
Collapse
|
10
|
Abstract
In this chapter, we review different imaging modalities, including radiography, computed tomography (CT), magnetic resonance imaging (MRI), ultrasound, and nuclear medicine scintigraphy, and their application to musculoskeletal neoplasm. Advantages and limitations of each modality are reviewed, and suggestions for imaging approach are provided.
Collapse
|