JCJulián Conejo-Mir

Julián Conejo-Mir

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Julián Conejo-Mir has a total of 48 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Melanoma Breslow Thickness Classification Using Ensemble-Based Knowledge Distillation With Semi-Supervised Convolutional Neural Networks

    by Juan P. Dominguez-Morales, Juan-Carlos Hernández-Rodríguez, Lourdes Duran-Lopez, Julián Conejo-Mir, Jose-Juan Pereyra-Rodriguez

    Melanoma is considered a global public health challenge and is responsible for more than 90% deaths related to skin cancer. Although the diagnosis of early melanoma is the main goal of dermoscopy, the discrimination between dermoscopic images of in situ and invasive melanomas can be a difficult task even for experienced dermatologists. Recent advances in artificial intelligence in the field of medical image analysis show that its application to dermoscopy with the aim of supporting and providing a second opinion to the medical expert could be of great interest. In this work, four datasets from different sources were used to train and evaluate deep learning models on in situ versus invasive melanoma classification and on Breslow thickness prediction. Supervised learning and semi-supervised learning using a multi-teacher ensemble knowledge distillation approach were considered and evaluated using a stratified 5-fold cross-validation scheme. The best models achieved AUCs of 0.6186 $\pm$ 0.0410 and of 0.7501 $\pm$ 0.0674 on the former and latter classification tasks, respectively. The best results were obtained using semi-supervised learning, with the best model achieving 0.7751 and 0.8164 AUC, respectively. The results obtained show that semi-supervised learning could improve the performance of trained models in different melanoma classification tasks compared to supervised learning. Automatic deep learning-based diagnosis systems could support medical professionals in their decision, serving as a second opinion or as a triage tool for medical centers.


  • Advances in treatment for lipoid proteinosis (Urbach–Wiethe disease): a case report and systematic review

    by Rocío C Bueno-Molina, Juan-Carlos Hernández-Rodríguez, Raquel Cabrera-Fuentes, Rocío Cabrera-Pérez, Julián Conejo-Mir Sánchez, José-Juan Pereyra-Rodríguez

    Background: Lipoid proteinosis (LP), also known as Urbach–Wiethe disease, is a rare autosomal recessive genodermatosis, caused by mutations in the ECM1 gene. This results in the deposition of periodic acid–Schiff (PAS)-positive, hyaline-like material on the skin, mucosae and internal organs. Objectives: To present a case report of LP and a systematic review to synthesize the scientific literature on the management of this uncommon and frequently missed diagnosis. Methods: We present a case report of a 48-year-old man with LP who exhibited significant improvement after oral acitretin therapy. To address the lack of large case–control studies on LP treatment, we performed a systematic review of the literature following the PRISMA 2020 criteria. The search was conducted in PubMed, Web of Science, Cochrane and Scopus databases from inception until June 2023. To assess the methodological quality of case reports and case series, we used the Joanna Briggs Collaboration critical appraisal tool. Results: We included 25 studies that met eligibility criteria. Data from 44 patients with a histopathologically confirmed diagnosis were analysed. Treatment ranged from systemic therapies (acitretin, etretinate, dimethyl sulfoxide, corticosteroids, penicillamine) to surgical or laser procedures. Regarding methodological quality, the main discrepancies arose in the reporting of participant characteristics and treatment interventions. Conclusions: Low-dose oral acitretin could have potential in managing LP, exhibiting fewer side-effects compared with other therapeutic agents. Further research is needed to establish more comprehensive and evidence-based treatment guidelines.


  • Open Access

    Prediction of melanoma Breslow thickness using deep transfer learning algorithms

    by Juan-Carlos Hernández-Rodríguez, Lourdes Durán-López, Juan P Domínguez-Morales, Juan Ortiz-Álvarez, Julián Conejo-Mir, Jose-Juan Pereyra-Rodriguez

    Background: The distinction between in situ melanoma (MIS) and invasive melanoma is challenging even for expert dermatologists. The use of pretrained convolutional neural networks (CNNs) as ancillary decision systems needs further research. Aim: To develop, validate and compare three deep transfer learning (DTL) algorithms to predict MIS vs. invasive melanoma and melanoma with a Breslow thickness (BT) of < 0.8 mm vs. ≥ 0.8 mm. Methods: A dataset of 1315 dermoscopic images of histopathologically confirmed melanomas was created from Virgen del Rocio University Hospital and open repositories of the International Skin Imaging Collaboration archive and Polesie S et al. (Dermatol Pract Concept 2021; 11:e2021079). The images were labelled as MIS or invasive melanoma and < 0.8 mm or ≥ 0.8 mm of BT. We conducted three trainings, and overall means for receiver operating characteristic (ROC) curves, sensitivity, specificity, positive and negative predictive value, and balanced diagnostic accuracy outcomes were evaluated on the test set with ResNetV2, EfficientNetB6 and InceptionV3. The results of 10 dermatologists were compared with the algorithms. Grad-CAM gradient maps were generated, highlighting relevant areas considered by the CNNs within the images. Results: EfficientNetB6 achieved the highest diagnostic accuracy for the comparison between MIS vs. invasive melanoma (61%) and BT < 0.8 mm vs. ≥ 0.8 mm (75%). For the BT comparison, ResNetV2 with an area under the ROC curve of 0.76 and InceptionV3 with an area under the ROC curve of 0.75, outperformed the results obtained by the dermatologist group with an area under the ROC curve of 0.70. Conclusion: EfficientNetB6 recorded the best prediction results, outperforming the dermatologists for the comparison of 0.8 mm of BT. DTL could be an ancillary aid to support dermatologists’ decisions in the near future.


  • Open Access

    Skin cancer mortality in Spain: adjusted mortality rates by province and related risk factors

    by Antonio Alcalá Ramírez del Puerto, Juan Carlos Hernández‐Rodriguez, Mercedes Sendín‐Martín, Juan Ortiz‐Alvarez, Julián Conejo‐Mir Sánchez, José Juan Pereyra‐Rodriguez

    Background: Ultraviolet radiation is the main environmental risk factor responsible for the development of skin cancer. Other occupational, socioeconomic, and environmental factors appear to be related to the risk of skin cancer. Furthermore, the factors appear to differ for melanoma and non‐melanoma skin cancer (NMSC). The purpose of this study is to analyze mortality rates of skin cancer in the different provinces of Spain and to determine the influence of socioeconomic conditions and other environmental and demographic factors in rates.Methods: Deaths from melanoma and NMSC in the period 2000–2019 were obtained as well as socioeconomic and environmental variables. Annual standardized mortality rates (SMR) were calculated for all Spanish provinces. The Pearson correlation coefficient was calculated.Results: The SMR of melanoma was 2.10/100,000 inhabitants, while that of NMSC was 1.28/100,000. At the provincial level, a great variability is confirmed. Gross domestic product showed a positive correlation with melanoma mortality but a negative correlation with NMSC. Other environmental and socioeconomic variables also showed correlation, as a positive correlation between tobacco sales and melanoma and between agricultural development and the NMSC.Conclusions: There are still important differences between each province that must be taken into account when planning health care and resource distribution. This ecological and province‐wise study helps to elucidate the relationship between social and ambient exposure determinants and skin cancer mortality in Spain.


  • Open Access

    Quantitative collagen analysis using second harmonic generation images for the detection of basal cell carcinoma with ex vivo multiphoton microscopy

    by Mercedes Sendín‐Martín, Jasmine Posner, Ucalene Harris, Matthew Moronta, Julián Conejo‐Mir Sánchez, Sushmita Mukherjee, Milind Rajadhyaksha, Kivanc Kose, Manu Jain

    Basal cell carcinoma (BCC) is the most common skin cancer, and its incidence is rising. Millions of benign biopsies are performed annually for BCC diagnosis, increasing morbidity, and healthcare costs. Non‐invasive in vivo technologies such as multiphoton microscopy (MPM) can aid in diagnosing BCC, reducing the need for biopsies. Furthermore, the second harmonic generation (SHG) signal generated from MPM can classify and prognosticate cancers based on extracellular matrix changes, especially collagen type I. We explored the potential of MPM to differentiate collagen changes associated with different BCC subtypes compared to normal skin structures and benign lesions. Quantitative analysis such as frequency band energy analysis in Fourier domain, CurveAlign and CT‐FIRE fibre analysis was performed on SHG images from 52 BCC and 12 benign lesions samples. Our results showed that collagen distribution is more aligned surrounding BCCs nests compared to the skin's normal structures (p < 0.001) and benign lesions (p < 0.001). Also, collagen was orientated more parallelly surrounding indolent BCC subtypes (superficial and nodular) versus those with more aggressive behaviour (infiltrative BCC) (p = 0.021). In conclusion, SHG signal from type I collagen can aid not only in the diagnosis of BCC but could be useful for prognosticating these tumors. Our initial results are limited to a small number of samples, requiring large‐scale studies to validate them. These findings represent the groundwork for future in vivo MPM for diagnosis and prognosis of BCC.


  • Open Access

    Dropout Rate in Digital Health Interventions for the Prevention of Skin Cancer: Systematic Review, Meta-analysis, and Metaregression

    by Juan-Carlos Hernández-Rodríguez, Cristina García-Muñoz, Juan Ortiz-Álvarez, Francesc Saigí-Rubió, Julián Conejo-Mir, Jose-Juan Pereyra-Rodriguez

    Background: Digital strategies are innovative approaches to the prevention of skin cancer, but the attrition following this kind of intervention needs to be analyzed. Objective: The aim of this paper is to assess the dropouts from studies focused on digital strategies for the prevention of skin cancer. Methods: We conducted this systematic review with meta-analyses and metaregression according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statements. Search terms for skin cancer, digital strategies, and prevention were combined to search PubMed, Scopus, Web of Science, CINAHL, and Cochrane Library from inception until July 2022. Randomized clinical trials that reported dropouts of participants and compared digital strategies with other interventions to prevent skin cancer in healthy or disease-free participants were included. Two independent reviewers extracted data for analysis. The Revised Cochrane Collaboration Bias tool was employed. We calculated the pooled dropout rate of participants through a meta-analysis of proportions and examined whether dropout was more or less frequent in digital interventions against comparators via an odds ratio (OR) meta-analysis. Data were pooled using a random-effects model. Subgroup meta-analyses were conducted in a meta-analysis of proportions and OR meta-analysis to assess the dropout events when data were sorted by digital interventions or control comparator. A univariate metaregression based on a random-effects model assessed possible moderators of dropout. Participants’ dropout rates as pooled proportions were calculated for all groups combined, and the digital and comparator groups separately. OR>1 indicated higher dropouts for digital-based interventions. Metaregressions were performed for age, sex, length of intervention, and sample size. Results: A total of 17 studies were included. The overall pooled dropout rate was 9.5% (95% CI 5.0-17.5). The subgroup meta-analysis of proportions revealed a dropout rate of 11.6% for digital strategies (95% CI 6.8-19.0) and 10.0% for comparators (95% CI 5.5-17.7). A trend of higher dropout rates for digital strategies was observed in the overall (OR 1.16, 95% CI 0.98-1.36) and subgroup OR meta-analysis, but no significant differences were found between the groups. None of the covariates moderated the effect size in the univariate metaregression. Conclusions: Digital strategies had a higher dropout rate compared to other prevention interventions, but the difference was not significant. Standardization is needed regarding reporting the number of and reasons for dropouts. Trial Registration: International Prospective Register of Systematic Reviews (PROSPERO) CRD42022329669; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=329669


  • Systemic sclerosis mortality trends in Spain from 1980 to 2019: age–period–cohort and joinpoint analysis

    by Juan-Carlos Hernández-Rodríguez, Mercedes Sendín-Martín, Antonio-José Durán-Romero, Juan Ortiz Álvarez, Julián Conejo-Mir, José-Juan Pereyra-Rodríguez

    Background: Systemic sclerosis (SSc) is an autoimmune chronic rheumatic disease with a high mortality rate, which continues to be a challenge for clinicians today. Aim: To assess changes in mortality trends in the Spanish SSc population between 1980 and 2019, taking into account the independent effects of sex, age, time period and birth cohort. Methods: SSc death records and mid-year population data were collected from the National Statistics Institute. Age-standardized mortality rates were calculated for the overall population and for each sex (male, female) and age group (5-year groups). Significant changes in mortality trends were identified by joinpoint regressions. An age–period–cohort (APC) analysis and potential years of life lost (PYLL) analysis were performed to identify the burden of SSc. Results: Age-standardized mortality rates due to SSc increased from 1.87 (95% CI 1.00–3.02) per 1 000 000 inhabitants between 1980 and 1984, to 2.47 (95% CI 1.74–3.02) per 1 000 000 inhabitants between 2015 and 2019. The relative risk of mortality fell in all groups in cohorts born after 1990. The PYLL rates showed a gradual rise for both sexes. Conclusion: There was an increase in overall SSc mortality in Spain during the 39 years evaluated, although there was a progressive drop for men.


  • Hidradenitis suppurativa and Down syndrome in a single‐centre sample: A cross‐sectional study

    by Juan‐Carlos Hernández‐Rodríguez, Guiovana‐Fernanda Osorio‐Gómez, Juan Ortiz‐Álvarez, José‐Antonio Lebrón‐Martín, Julián Conejo‐Mir, José‐Juan Pereyra‐Rodríguez

    Down syndrome (DS) has been related to a higher risk of hidradenitis suppurativa (HS). This cross‐sectional study assessed DS patients with HS in a Spanish single‐centre sample. DS participants presented a lower age of onset, age at diagnosis and time to diagnosis. Also, DS was not associated with the severity of HS measured by baseline IHS4.


  • Open Access

    Novel Non-Invasive Quantification and Imaging of Eumelanin and DHICA Subunit in Skin Lesions by Raman Spectroscopy and MCR Algorithm: Improving Dysplastic Nevi Diagnosis

    by José Javier Ruiz, Monica Marro, Ismael Galván, José Bernabeu-Wittel, Julián Conejo-Mir, Teresa Zulueta-Dorado, Ana Belén Guisado-Gil, Pablo Loza-Álvarez

    Malignant melanoma (MM) is the most aggressive form of skin cancer, and around 30% of them may develop from pre-existing dysplastic nevi (DN). Diagnosis of DN is a relevant clinical challenge, as these are intermediate lesions between benign and malignant tumors, and, up to date, few studies have focused on their diagnosis. In this study, the accuracy of Raman spectroscopy (RS) is assessed, together with multivariate analysis (MA), to classify 44 biopsies of MM, DN and compound nevus (CN) tumors. For this, we implement a novel methodology to non-invasively quantify and localize the eumelanin pigment, considered as a tumoral biomarker, by means of RS imaging coupled with the Multivariate Curve Resolution-Alternative Least Squares (MCR-ALS) algorithm. This represents a step forward with respect to the currently established technique for melanin analysis, High-Performance Liquid Chromatography (HPLC), which is invasive and cannot provide information about the spatial distribution of molecules. For the first time, we show that the 5, 6-dihydroxyindole (DHI) to 5,6-dihydroxyindole-2-carboxylic acid (DHICA) ratio is higher in DN than in MM and CN lesions. These differences in chemical composition are used by the Partial Least Squares-Discriminant Analysis (PLS-DA) algorithm to identify DN lesions in an efficient, non-invasive, fast, objective and cost-effective method, with sensitivity and specificity of 100% and 94.1%, respectively.


  • Classification of Basal Cell Carcinoma in Ex Vivo Confocal Microscopy Images from Freshly Excised Tissues Using a Deep Learning Algorithm

    by Mercedes Sendín-Martín, Manuel Lara-Caro, Ucalene Harris, Matthew Moronta, Anthony Rossi, Erica Lee, Chih-Shan Jason Chen, Kishwer Nehal, Julián Conejo-Mir Sánchez, José-Juan Pereyra-Rodríguez, et al.