Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging
Abstract Background Neuroblastoma is one common pediatric malignancy notorious for high temporal and spatial heterogeneities. More than half of its patients develop distant metastases involving vascularized organs, especially the bone marrow. It is thus necessary to have an economical, noninvasive m...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2023-10-01
|
| Series: | Cancer Innovation |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/cai2.92 |
| _version_ | 1827779360895008768 |
|---|---|
| author | Lin Lv Zhengtao Zhang Dongbo Zhang Qinchang Chen Yuanfang Liu Ya Qiu Wen Fu Xuntao Yin Xiong Chen |
| author_facet | Lin Lv Zhengtao Zhang Dongbo Zhang Qinchang Chen Yuanfang Liu Ya Qiu Wen Fu Xuntao Yin Xiong Chen |
| author_sort | Lin Lv |
| collection | DOAJ |
| description | Abstract Background Neuroblastoma is one common pediatric malignancy notorious for high temporal and spatial heterogeneities. More than half of its patients develop distant metastases involving vascularized organs, especially the bone marrow. It is thus necessary to have an economical, noninvasive method without much radiation for follow‐ups. Radiomics has been used in many cancers to assist accurate diagnosis but not yet in bone marrow metastasis in neuroblastoma. Methods A total of 182 patients with neuroblastoma were retrospectively collected and randomly divided into the training and validation sets. Five‐hundred and seventy‐two radiomics features were extracted from magnetic resonance imaging, among which 41 significant ones were selected via T‐test for model development. We attempted 13 machine‐learning algorithms and eventually chose three best‐performed models. The integrative performance evaluations are based on the area under the curves (AUCs), calibration curves, risk deciles plots, and other indexes. Results Extreme gradient boosting, random forest (RF), and adaptive boosting were the top three to predict bone marrow metastases in neuroblastoma while RF was the most accurate one. Its AUC was 0.90 (0.86–0.93), F1 score was 0.82, sensitivity was 0.76, and negative predictive value was 0.79 in the training set. The values were 0.82 (0.71–0.93), 0.80, 0.75, and 0.92 in the validation set, respectively. Conclusions Radiomics models are likely to contribute more to metastatic diagnoses and the formulation of personalized healthcare strategies in clinics. It has great potential of being a revolutionary method to replace traditional interventions in the future. |
| first_indexed | 2024-03-11T14:47:51Z |
| format | Article |
| id | doaj.art-b766cc8bc0834971868abf6a579049bb |
| institution | Directory Open Access Journal |
| issn | 2770-9183 |
| language | English |
| last_indexed | 2024-03-11T14:47:51Z |
| publishDate | 2023-10-01 |
| publisher | Wiley |
| record_format | Article |
| series | Cancer Innovation |
| spelling | doaj.art-b766cc8bc0834971868abf6a579049bb2023-10-30T10:36:07ZengWileyCancer Innovation2770-91832023-10-012540541510.1002/cai2.92Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imagingLin Lv0Zhengtao Zhang1Dongbo Zhang2Qinchang Chen3Yuanfang Liu4Ya Qiu5Wen Fu6Xuntao Yin7Xiong Chen8Department of Urology Surgery SunYat‐Sen Memorial Hospital Guangzhou Guangdong ChinaGuangzhou Women and Children's Medical Center Guangzhou Guangdong ChinaBreast Tumor Center Sun Yat‐Sen Memorial Hospital Guangzhou Guangdong ChinaGuangdong Provincial People's Hospital Guangzhou Guangdong ChinaDepartment of Radiology Sun Yat‐Sen Memorial Hospital Guangzhou Guangdong ChinaDepartment of Radiology Sun Yat‐Sen Memorial Hospital Guangzhou Guangdong ChinaGuangzhou Women and Children's Medical Center Guangzhou Guangdong ChinaDepartment of Radiology Guangzhou Women and Children's Medical Center Guangzhou Guangdong ChinaDepartment of Urology Surgery SunYat‐Sen Memorial Hospital Guangzhou Guangdong ChinaAbstract Background Neuroblastoma is one common pediatric malignancy notorious for high temporal and spatial heterogeneities. More than half of its patients develop distant metastases involving vascularized organs, especially the bone marrow. It is thus necessary to have an economical, noninvasive method without much radiation for follow‐ups. Radiomics has been used in many cancers to assist accurate diagnosis but not yet in bone marrow metastasis in neuroblastoma. Methods A total of 182 patients with neuroblastoma were retrospectively collected and randomly divided into the training and validation sets. Five‐hundred and seventy‐two radiomics features were extracted from magnetic resonance imaging, among which 41 significant ones were selected via T‐test for model development. We attempted 13 machine‐learning algorithms and eventually chose three best‐performed models. The integrative performance evaluations are based on the area under the curves (AUCs), calibration curves, risk deciles plots, and other indexes. Results Extreme gradient boosting, random forest (RF), and adaptive boosting were the top three to predict bone marrow metastases in neuroblastoma while RF was the most accurate one. Its AUC was 0.90 (0.86–0.93), F1 score was 0.82, sensitivity was 0.76, and negative predictive value was 0.79 in the training set. The values were 0.82 (0.71–0.93), 0.80, 0.75, and 0.92 in the validation set, respectively. Conclusions Radiomics models are likely to contribute more to metastatic diagnoses and the formulation of personalized healthcare strategies in clinics. It has great potential of being a revolutionary method to replace traditional interventions in the future.https://doi.org/10.1002/cai2.92bone marrow metastasismachine learningmagnetic resonance imagingneuroblastomaradiomics |
| spellingShingle | Lin Lv Zhengtao Zhang Dongbo Zhang Qinchang Chen Yuanfang Liu Ya Qiu Wen Fu Xuntao Yin Xiong Chen Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging Cancer Innovation bone marrow metastasis machine learning magnetic resonance imaging neuroblastoma radiomics |
| title | Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging |
| title_full | Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging |
| title_fullStr | Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging |
| title_full_unstemmed | Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging |
| title_short | Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging |
| title_sort | machine learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging |
| topic | bone marrow metastasis machine learning magnetic resonance imaging neuroblastoma radiomics |
| url | https://doi.org/10.1002/cai2.92 |
| work_keys_str_mv | AT linlv machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging AT zhengtaozhang machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging AT dongbozhang machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging AT qinchangchen machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging AT yuanfangliu machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging AT yaqiu machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging AT wenfu machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging AT xuntaoyin machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging AT xiongchen machinelearningradiomicstopredictbonemarrowmetastasisofneuroblastomausingmagneticresonanceimaging |