Artificial intelligence performance in detecting lymphoma from medical imaging: a systematic review and meta-analysis.

BACKGROUND: Accurate diagnosis and early treatment are essential in the fight against lymphatic cancer. The application of artificial intelligence (AI) in the field of medical imaging shows great potential, but the diagnostic accuracy of lymphoma is unclear. This study was done to systematically review and meta-analyse researches concerning the diagnostic performance of AI in detecting lymphoma using medical imaging for the first time. METHODS: Searches were conducted in Medline, Embase, IEEE and Cochrane up to December 2023. Data extraction and assessment of the included study quality were independently conducted by two investigators. Studies that reported the diagnostic performance of an AI model/s for the early detection of lymphoma using medical imaging were included in the systemic review. We extracted the binary diagnostic accuracy data to obtain the outcomes of interest: sensitivity (SE), specificity (SP), and Area Under the Curve (AUC). The study was registered with the PROSPERO, CRD42022383386. RESULTS: Thirty studies were included in the systematic review, sixteen of which were meta-analyzed with a pooled sensitivity of 87% (95%CI 83-91%), specificity of 94% (92-96%), and AUC of 97% (95-98%). Satisfactory diagnostic performance was observed in subgroup analyses based on algorithms types (machine learning versus deep learning, and whether transfer learning was applied), sample size (≤ 200 or >  200), clinicians versus AI models and geographical distribution of institutions (Asia versus non-Asia). CONCLUSIONS: Even if possible overestimation and further studies with a better standards for application of AI algorithms in lymphoma detection are needed, we suggest the AI may be useful in lymphoma diagnosis.

Pregnancy Loss in Relation to the Risks of Female-Specific Cancers in a Population-Based Cohort and Mendelian Randomization Study - China, 2004-2017.

What is already known about this topic?: Limited evidence exists regarding the relationship between pregnancy loss and female-specific cancers within the Chinese population from prospective cohort studies. What is added by this report?: Terminations were associated with a 13% lower risk of endometrial cancer, whereas stillbirths were related to an 18% higher risk of cervical cancer. Rural residents with a history of pregnancy loss experienced a 19% and 38% increased risk of breast and cervical cancers, respectively, compared to their urban counterparts. Moreover, a positive graded relationship between live births and pregnancy loss on cervical cancer was observed. What are the implications for public health practice?: This study has significant implications for identifying women at an increased risk for breast and genital cancers and contributes to the development of effective public health strategies for female cancer prevention. Future research on reproductive history, particularly in rural areas, should be given priority in efforts to improve female cancer screening and early detection.

Correction: Unassisted Clinicians Versus Deep Learning-Assisted Clinicians in Image-Based Cancer Diagnostics: Systematic Review With Meta-analysis.

[This corrects the article DOI: 10.2196/43832.].

Unassisted Clinicians Versus Deep Learning-Assisted Clinicians in Image-Based Cancer Diagnostics: Systematic Review With Meta-analysis.

BACKGROUND: A number of publications have demonstrated that deep learning (DL) algorithms matched or outperformed clinicians in image-based cancer diagnostics, but these algorithms are frequently considered as opponents rather than partners. Despite the clinicians-in-the-loop DL approach having great potential, no study has systematically quantified the diagnostic accuracy of clinicians with and without the assistance of DL in image-based cancer identification. OBJECTIVE: We systematically quantified the diagnostic accuracy of clinicians with and without the assistance of DL in image-based cancer identification. METHODS: PubMed, Embase, IEEEXplore, and the Cochrane Library were searched for studies published between January 1, 2012, and December 7, 2021. Any type of study design was permitted that focused on comparing unassisted clinicians and DL-assisted clinicians in cancer identification using medical imaging. Studies using medical waveform-data graphics material and those investigating image segmentation rather than classification were excluded. Studies providing binary diagnostic accuracy data and contingency tables were included for further meta-analysis. Two subgroups were defined and analyzed, including cancer type and imaging modality. RESULTS: In total, 9796 studies were identified, of which 48 were deemed eligible for systematic review. Twenty-five of these studies made comparisons between unassisted clinicians and DL-assisted clinicians and provided sufficient data for statistical synthesis. We found a pooled sensitivity of 83% (95% CI 80%-86%) for unassisted clinicians and 88% (95% CI 86%-90%) for DL-assisted clinicians. Pooled specificity was 86% (95% CI 83%-88%) for unassisted clinicians and 88% (95% CI 85%-90%) for DL-assisted clinicians. The pooled sensitivity and specificity values for DL-assisted clinicians were higher than for unassisted clinicians, at ratios of 1.07 (95% CI 1.05-1.09) and 1.03 (95% CI 1.02-1.05), respectively. Similar diagnostic performance by DL-assisted clinicians was also observed across the predefined subgroups. CONCLUSIONS: The diagnostic performance of DL-assisted clinicians appears better than unassisted clinicians in image-based cancer identification. However, caution should be exercised, because the evidence provided in the reviewed studies does not cover all the minutiae involved in real-world clinical practice. Combining qualitative insights from clinical practice with data-science approaches may improve DL-assisted practice, although further research is required. TRIAL REGISTRATION: PROSPERO CRD42021281372; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=281372.

Deep learning in image-based breast and cervical cancer detection: a systematic review and meta-analysis.

Accurate early detection of breast and cervical cancer is vital for treatment success. Here, we conduct a meta-analysis to assess the diagnostic performance of deep learning (DL) algorithms for early breast and cervical cancer identification. Four subgroups are also investigated: cancer type (breast or cervical), validation type (internal or external), imaging modalities (mammography, ultrasound, cytology, or colposcopy), and DL algorithms versus clinicians. Thirty-five studies are deemed eligible for systematic review, 20 of which are meta-analyzed, with a pooled sensitivity of 88% (95% CI 85-90%), specificity of 84% (79-87%), and AUC of 0.92 (0.90-0.94). Acceptable diagnostic performance with analogous DL algorithms was highlighted across all subgroups. Therefore, DL algorithms could be useful for detecting breast and cervical cancer using medical imaging, having equivalent performance to human clinicians. However, this tentative assertion is based on studies with relatively poor designs and reporting, which likely caused bias and overestimated algorithm performance. Evidence-based, standardized guidelines around study methods and reporting are required to improve the quality of DL research.

Clinical risk factors associated with recurrence of ischemic stroke within two years: A cohort study.

To identify the clinical risk factors and investigate the efficacy of a classification model based on the identified factors for predicting 2-year recurrence after ischemic stroke.From June 2017 to January 2019, 358 patients with first-ever ischemic stroke were enrolled and followed up in Shenzhen Traditional Chinese Medicine Hospital. Demographic and clinical characteristics were recorded by trained medical staff. The outcome was defined as recurrence within 2 years. A multivariate logistic regression model with risk factors and their interaction effects was established and evaluated.The mean (standard deviation) age of the participants was 61.6 (12.1) years, and 101 (28.2%) of the 358 patients were female. The common comorbidities included hypertension (286 patients, 79.9%), diabetes (148 patients, 41.3%), and hyperlipidemia (149 patients, 41.6%). The 2-year recurrence rate was 30.7%. Of the 23 potential risk factors, 10 were significantly different between recurrent and non-recurrent subjects in the univariate analysis. A multivariate logistic regression model was developed based on 10 risk factors. The significant variables include diabetes mellitus, smoking status, peripheral artery disease, hypercoagulable state, depression, 24 h minimum systolic blood pressure, 24 h maximum diastolic blood pressure, age, family history of stroke, NIHSS score status. The area under the receiver operating characteristic curve (ROC) was 0.78 (95% confidence interval: 0.726-0.829) with a sensitivity of 0.61 and a specificity of 0.81, indicating a potential predictive ability.Ten risk factors were identified, and an effective classification model was built. This may aid clinicians in identifying high-risk patients who would benefit most from intensive follow-up and aggressive risk factor reduction.The clinical trial registration number: ChiCTR1800019647.

Exposure to tobacco smoke and stillbirth: a national prospective cohort study in rural China.

INTRODUCTION: Parental exposure to tobacco smoke has been associated with an increased risk of stillbirth, while only a few studies took the overall parental tobacco exposure status into consideration. We aim to explore the relationship of parental smoking and passive smoking before and during pregnancy with stillbirth in a large Chinese rural cohort. DESIGN: 248 501 couples were enrolled in a national prospective cohort study conducted in rural China. Parental exposure to tobacco smoke before and during pregnancy, along with other risk factors, was ascertained by questionnaires. Pregnancy outcomes were recorded by physicians. RESULTS: The ORs (Odds Ratios) of maternal active smoking, maternal passive smoking, paternal active smoking and paternal passive smoking were 2.07 (95% CI 1.25 to 3.41), 1.22 (95% CI 1.01 to 1.47), 1.36 (95% CI 1.13 to 1.63) and 1.10 (95% CI 0.87 to 1.39), respectively. The rates of stillbirth increased from 0.31% for the maternal non-smoking group to 0.43% for the smoking cessation during pregnancy group, to 0.64% for the decreased smoking group and 1.28% for the continuing smoking group. A similar pattern was found in the change in paternal smoking status and stillbirth. Stratified by maternal passive smoking, the OR of paternal smoking was 1.35 (95% CI 1.13 to 1.61) in the maternal non-smoking group and 1.67 (95% CI 1.09 to 2.56) in the maternal passive smoking group. CONCLUSIONS: Parental exposure to tobacco smoke increased the risk of stillbirth, especially for those continuing smoking during pregnancy. Paternal smoking is an independent risk factor for stillbirth despite maternal passive smoking status.

Risk factors of stillbirth in rural China: A national cohort study.

People living in rural China are more frequently exposed to some specific risk factors which made stillbirth rate higher than urban areas. National Free Preconception Health Examination Project was launched to investigate these risk factors and collected a representative sample of 248501 participants from 31 provinces in China from 2010 to 2013. Parental risk factors were ascertained twice before and during pregnancy respectively by questionnaires. Stillbirth or live birth were recorded by trained physicians. In the analysis, nested case-control study was conducted, and propensity score matching method was used to adjust the confounding. Multi-level logistic regression was used to fit for multi-level sampling. The overall stillbirth rate was 0.35% in rural China, it was higher in North (0.42%) and West (0.64%) areas. Maternal exposure to pesticide (OR (95%CI 1.06, 3.39)), hypertension (OR = 1.58 (95%CI 1.07, 2.34)), lack of appetite for vegetables (OR = 1.99 (95%CI 1.00, 3.93)), stress (compared with no pressure, OR of a little pressure was 1.34(95% CI 1.02, 1.76)); paternal exposure to smoking (OR = 1.22 (95% CI 1.02, 1.46)), organic solvents (OR = 1.64 (95% CI 1.01, 2.69)) were found independent risk factors of stillbirth. Folacin intake 3 months before pregnancy (OR = 0.72 (95%) CI 0.59, 0.89), folacin intake 1-2 months before pregnancy (OR = 0.71 (95% CI 0.55, 0.92)), folacin intake after pregnancy (OR = 0.81 (95% CI 0.65, 1.02) for) were protect factors of stillbirth. Maternal pesticide exposure, lack of vegetables, stress, paternal smoking and exposure to organic solvents were risk factors of stillbirth. Folic acid intake was protective factor of stillbirth, no matter when the intake began.

Both maternal and paternal risk factors for term singleton low birthweight infants in rural Chinese population: a population-based, retrospective cohort study.

No large population-based study has focused on both maternal paternal risk factors for low birthweight (LBW) in China. We aimed to identify parental risk factors associated with LBW.A population-based, retrospective cohort study was conducted on 202,725 singleton infants at 37-42 weeks. These term singleton newborns were classified as LBW with birthweight ≤2500 g(TLBW) and normal birthweight between 50th to 97th percentile (TNBW 50th-97th) according to Chinese singleton norms. Multiple logistic regression analyses were used to find those parental risk factors of LBW by comparing two groups. TLBW and TNBW(50th-97th) occupied 4.8% and 70.8% of the study population, respectively. Logistic regression showed a significant association with positive maternal hepatitis B surface antigen (RR = 1.979, P = 0.047), irregular folic acid intake (RR = 1.152, P = 0.003), paternal history of varicocele (RR = 2.404, P = 0.003) and female babies (RR = 1.072, P = 0.046). Maternal smoking, hypertension and history of stillbirth were found related to LBW but no statistically significant. Positive maternal hepatitis B surface antigen, irregular folic acid intake, paternal history of varicocele had a negative effect on birth weight. Measures are necessarily taken to avoid them to improve pregnancy outcomes. Further studies should be done to investigate each detailed risk factors on LBW.

A cocktail of p16INK4a and Ki-67, p16INK4a and minichromosome maintenance protein 2 as triage tests for human papillomavirus primary cervical cancer screening.

Most human papillomavirus (HPV) infections are transient and additional triage approaches should be built after HPV-based primary cervical cancer screening. We evaluated the accuracy of p16/Ki-67 and p16/mcm2 dual staining as biomarkers for triaging HPV positive women in China. 4070 participants aged 35 to 64 years attending ongoing cervical cancer screening were enrolled in 2015-2016. Cervical exfoliated cells were collected for HPV DNA analysis and the residual positive specimens were tested for liquid-based cytology and biomarkers. Women infected with HPV 16/18 type or other 12 high-risk HPV types with abnormal cytology results received colposcopy. We found the positive rates of both biomarkers increased significantly with histology severity. p16/Ki-67 positivity in HPV16/18 group, other 12 high-risk HPV group and HPV negative group was 50.0%, 33.7% and 8.9%, respectively. The corresponding p16/mcm2 positivity was 70.0%, 56.3% and 6.7%, respectively. The sensitivity and specificity of p16/Ki-67 for CIN2+ in all HPV-positive women were 91.7% and 63.5%, with a referral rate of 36.2%, while p16/mcm2 were 87.5% and 42.1%, with a referral rate of 58.4%, respectively. The sensitivity of p16/Ki-67 increased to 95.8% for CIN2+ and 100% for CIN3+ when combined with high-grade cytology, without decrease in specificity. Our studies suggest that p16/Ki-67 is an efficient triaging biomarker for HPV-positive women and could reduce colposcopy workload. p16/mcm2 is more sensitive compared with cytology for identifying cervical lesions.

Prognostic analysis for Chinese patients with stage I ovarian endometrioid carcinoma.

BACKGROUND: This study aimed to identify the clinical and pathological characteristics and the possible prognostic factors for Chinese patients with early-stage ovarian endometrioid carcinoma. METHODS: The present study reviewed the medical records of patients who received initial treatment and a postoperative pathological diagnosis of ovarian endometrioid carcinoma at our center. In all, 78 patients had stage I ovarian endometrioid carcinoma. RESULTS: In this series, the 5-year overall survival rate and 5-year disease-free survival (DFS) rates of patients with stage I ovarian endometrioid carcinoma was 98.7% and 87.2%, respectively. Univariate analysis showed the factors that influence the DFS rates include menopausal status, FIGO stage, histological grade, lymphadenectomy, cytology of ascites. Multivariate analysis showed that grade 3 and lymphadenectomy were the independent prognostic factors of DFS for Stage I ovarian endometrioid carcinoma (P = 0.0259, 0.0276 respectively). However, the coexisting endometriosis, concomitant endometrial disorders, dissection of para-aortic lymph node and more courses of thermotherapy had no influence on DFS. Besides, it was found that 19.3% of patients in this series had synchronous early stage and well-to-moderate differentiated endometrial carcinoma. CONCLUSIONS: Grade 3 and lymphadenectomy were indicated as the independent factors of DFS for stage I patients with ovarian endometrioid carcinoma. The endometrial changes should be considered seriously when fertility-sparing surgery was planned.

Serum levels of leptin, adiponectin and resistin in relation to clinical characteristics in normal pregnancy and preeclampsia.

BACKGROUND: Alterations in serum adipokines in preeclampsia remain vague. We investigated the roles of leptin, adiponectin and resistin and their relationships with clinical characteristics in normotensive and preeclamptic patients. METHODS: A case-control study was carried out in a cohort of 74 preeclampsia(PE) and 79 healthy pregnant women. Serum levels of leptin, adiponectin and resistin were measured by enzyme-linked immunosorbent assay. RESULTS: The mean body mass index(BMI), the serum leptin and resistin levels were significantly higher in the PE group than in the control group (p<0.001). The resistin/creatinine ratio was also higher in the PE group than in the control group (p=0.018). No significant difference was observed in the serum adiponectin level between both groups. Serum leptin levels were positively correlated with BMI (r=0.301, p<0.001) and negatively correlated with newborn birth weight (r=-0.435, p<0.001). Serum resistin levels were also negatively correlated with birth weight (r=-0.229, p=0.005) but were unrelated to BMI. Logistic regression showed that BMI≥28 and serum leptin levels were independent factors of PE. Leptin was a potential mediator between BMI and PE (p<0.001), and the mediating effect accounted for 22.54% of the total effect. CONCLUSIONS: Leptin, resistin, and obesity play important roles in the onset of PE. Leptin and resistin may have some impacts on the fetal growth and development.