ABSTRACT
BACKGROUND: The diagnosis of Parkinson's disease is currently based on clinical evaluation. Despite clinical hallmarks, unfortunately, the error rate is still significant. Low in-vivo diagnostic accuracy of clinical evaluation mainly relies on the lack of quantitative biomarkers for an objective motor performance assessment. Non-invasive technologies, such as wearable sensors, coupled with machine learning algorithms, assess quantitatively and objectively the motor performances, with possible benefits either for in-clinic and at-home settings. We conducted a systematic review of the literature on machine learning algorithms embedded in smart devices in Parkinson's disease diagnosis. METHODS: Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, we searched PubMed for articles published between December, 2007 and July, 2023, using a search string combining "Parkinson's disease" AND ("healthy" or "control") AND "diagnosis", within the Groups and Outcome domains. Additional search terms included "Algorithm", "Technology" and "Performance". RESULTS: From 89 identified studies, 47 met the inclusion criteria based on the search string and four additional studies were included based on the Authors' expertise. Gait emerged as the most common parameter analysed by machine learning models, with Support Vector Machines as the prevalent algorithm. The results suggest promising accuracy with complex algorithms like Random Forest, Support Vector Machines, and K-Nearest Neighbours. DISCUSSION: Despite the promise shown by machine learning algorithms, real-world applications may still face limitations. This review suggests that integrating machine learning with wearable sensors has the potential to improve Parkinson's disease diagnosis. These tools could provide clinicians with objective data, potentially aiding in earlier detection.
ABSTRACT
INTRODUCTION: Epidemiological studies have consistently shown an inverse association between cigarette smoking and Parkinson's disease. Literature indicates that both current and former smokers have a reduced risk of developing PD compared to non-smokers. If smoking protects against Parkinson's disease risk or, conversely, smoking habit is abated due to the disease itself, according to the reverse causation, is still an unsolved question. METHODS: 118 patients from the UK Brain Bank with an alive clinical diagnosis of Parkinson's disease were enrolled. Post-mortem validation served as the gold standard for diagnosis to divide the population into true positive and false positive groups. Patient charts were reviewed to extract smoking exposure information and statistical analyses were conducted to determine the odds associated with smoking in the two diagnostic groups. RESULTS: Among alive clinically diagnosed patients with Parkinson's disease, 53 % had no smoking exposure. In the True Positive group, 58 % had no smoking exposure, while this proportion was lower in the False Positive group at 46 %. The Odds Ratio for the association between smoking exposure and the two groups was 0.63 (95 % CI: 0.32-1.37). The Chi-square test yielded a p-value of 0.2804. CONCLUSIONS: Our findings emphasize the role of smoking exposure in Parkinson's diagnosis. The results indicate that the observed association is not specific to idiopathic Parkinson's disease but rather a broader phenomenon encompassing various parkinsonian disorders. This suggests a potential common neuroprotective effect of smoking, shared risk factors, or supports the reverse causation hypothesis where parkinsonian symptoms reduce smoking exposure.
Subject(s)
Parkinson Disease , Smoking , Humans , Parkinson Disease/epidemiology , United Kingdom/epidemiology , Female , Male , Aged , Case-Control Studies , Middle Aged , Smoking/adverse effects , Smoking/epidemiology , Tissue Banks , Aged, 80 and overABSTRACT
Stroke, a complex and heterogeneous disease, is a leading cause of morbidity and mortality worldwide. The timely therapeutic intervention significantly impacts patient outcomes, but early stroke diagnosis is challenging due to the lack of specific diagnostic biomarkers. This review critically examines the literature for potential biomarkers that may aid in early diagnosis, differentiation between ischemic and hemorrhagic stroke, and prediction of hemorrhagic transformation in ischemic stroke. After a thorough analysis, four promising biomarkers were identified: Antithrombin III (ATIII), fibrinogen, and ischemia-modified albumin (IMA) for diagnostic purposes; glial fibrillary acidic protein (GFAP), micro RNA 124-3p, and a panel of 11 metabolites for distinguishing between ischemic and hemorrhagic stroke; and matrix metalloproteinase-9 (MMP-9), s100b, and interleukin 33 for predicting hemorrhagic transformation. We propose a biomarker panel integrating these markers, each reflecting different pathophysiological stages of stroke, that could significantly improve stroke patients' early detection and treatment. Despite promising results, further research and validation are needed to demonstrate the clinical utility of this proposed panel for routine stroke treatment.
Subject(s)
Brain Ischemia , Hemorrhagic Stroke , Stroke , Humans , Brain Ischemia/diagnosis , Biomarkers , Serum Albumin , Stroke/diagnosisABSTRACT
Since its first introduction, levodopa has become the cornerstone for the treatment of Parkinson's disease and remains the leading therapeutic choice for motor control therapy so far. Unfortunately, the subsequent appearance of abnormal involuntary movements, known as dyskinesias, is a frequent drawback. Despite the deep knowledge of this complication, in terms of clinical phenomenology and the temporal relationship during a levodopa regimen, less is clear about the pathophysiological mechanisms underpinning it. As the disease progresses, specific oscillatory activities of both motor cortical and basal ganglia neurons and variation in levodopa metabolism, in terms of the dopamine receptor stimulation pattern and turnover rate, underlie dyskinesia onset. This review aims to provide a global overview on levodopa-induced dyskinesias, focusing on pathophysiology, clinical manifestations, therapy management strategies and future directions.
ABSTRACT
INTRODUCTION: The present study explores brain connectivity in Parkinson's disease (PD) and in age matched healthy controls (HC), using quantitative EEG analysis, at rest and during a motor tasks. We also evaluated the diagnostic performance of the phase locking value (PLV), a measure of functional connectivity, in differentiating PD patients from HCs. METHODS: High-density, 64-channels, EEG data from 26 PD patients and 13 HC were analyzed. EEG signals were recorded at rest and during a motor task. Phase locking value (PLV), as a measure of functional connectivity, was evaluated for each group in a resting state and during a motor task for the following frequency bands: (i) delta: 2-4 Hz; (ii) theta: 5-7 Hz; (iii) alpha: 8-12 Hz; beta: 13-29 Hz; and gamma: 30-60 Hz. The diagnostic performance in PD vs. HC discrimination was evaluated. RESULTS: Results showed no significant differences in PLV connectivity between the two groups during the resting state, but a higher PLV connectivity in the delta band during the motor task, in HC compared to PD. Comparing the resting state versus the motor task for each group, only HCs showed a higher PLV connectivity in the delta band during motor task. A ROC curve analysis for HC vs. PD discrimination, showed an area under the ROC curve (AUC) of 0.75, a sensitivity of 100%, and a negative predictive value (NPV) of 100%. CONCLUSIONS: The present study evaluated the brain connectivity through quantitative EEG analysis in Parkinson's disease versus healthy controls, showing a higher PLV connectivity in the delta band during the motor task, in HC compared to PD. This neurophysiology biomarkers showed the potentiality to be explored in future studies as a potential screening biomarker for PD patients.
ABSTRACT
BACKGROUND: Directional deep brain stimulation (DBS) leads allow a fine-tuning control of the stimulation field, however, this new technology could increase the DBS programming time because of the higher number of the possible combinations used in directional DBS than in standard nondirectional electrodes. Neuroimaging leads localization techniques and local field potentials (LFPs) recorded from DBS electrodes implanted in basal ganglia are among the most studied biomarkers for DBS programing. OBJECTIVE: This study aimed to evaluate whether intraoperative LFPs beta power and neuroimaging reconstructions correlate with contact selection in clinical programming of DBS in patients with Parkinson disease (PD). MATERIALS AND METHODS: In this retrospective study, routine intraoperative LFPs recorded from all contacts in the subthalamic nucleus (STN) of 14 patients with PD were analyzed to calculate the beta band power for each contact. Neuroimaging reconstruction obtained through Brainlab Elements Planning software detected contacts localized within the STN. Clinical DBS programming contact scheme data were collected after one year from the implant. Statistical analysis evaluated the diagnostic performance of LFPs beta band power and neuroimaging data for identification of the contacts selected with clinical programming. We evaluated whether the most effective contacts identified based on the clinical response after one year from implant were also those with the highest level of beta activity and localized within the STN in neuroimaging reconstruction. RESULTS: LFPs beta power showed a sensitivity of 67%, a negative predictive value (NPV) of 84%, a diagnostic odds ratio (DOR) of 2.7 in predicting the most effective contacts as evaluated through the clinical response. Neuroimaging reconstructions showed a sensitivity of 62%, a NPV of 77%, a DOR of 1.20 for contact effectivity prediction. The combined use of the two methods showed a sensitivity of 87%, a NPV of 87%, a DOR of 2.7 for predicting the clinically more effective contacts. CONCLUSIONS: The combined use of LFPs beta power and neuroimaging localization and segmentations predict which are the most effective contacts as selected on the basis of clinical programming after one year from implant of DBS. The use of predictors in contact selection could guide clinical programming and reduce time needed for it.
Subject(s)
Deep Brain Stimulation , Parkinson Disease , Subthalamic Nucleus , Humans , Parkinson Disease/therapy , Parkinson Disease/surgery , Retrospective Studies , Deep Brain Stimulation/methods , Subthalamic Nucleus/diagnostic imaging , Subthalamic Nucleus/surgery , Subthalamic Nucleus/physiology , NeuroimagingABSTRACT
Movement disorders include a wide and heterogeneous variety of signs and syndromes, which are classified as hyperkinetic [...].
ABSTRACT
Quantitative indoor monitoring, in a low-invasive and accurate way, is still an unmet need in clinical practice. Indoor environments are more challenging than outdoor environments, and are where patients experience difficulty in performing activities of daily living (ADLs). In line with the recent trends of telemedicine, there is an ongoing positive impulse in moving medical assistance and management from hospitals to home settings. Different technologies have been proposed for indoor monitoring over the past decades, with different degrees of invasiveness, complexity, and capabilities in full-body monitoring. The major classes of devices proposed are inertial-based sensors (IMU), vision-based devices, and geomagnetic and radiofrequency (RF) based sensors. In recent years, among all available technologies, there has been an increasing interest in using RF-based technology because it can provide a more accurate and reliable method of tracking patients' movements compared to other methods, such as camera-based systems or wearable sensors. Indeed, RF technology compared to the other two techniques has higher compliance, low energy consumption, does not need to be worn, is less susceptible to noise, is not affected by lighting or other physical obstacles, has a high temporal resolution without a limited angle of view, and fewer privacy issues. The aim of the present narrative review was to describe the potential applications of RF-based indoor monitoring techniques and highlight their differences compared to other monitoring technologies.
Subject(s)
Gait Analysis , Telemedicine , Humans , Tremor , Activities of Daily Living , Polysomnography , Vital SignsABSTRACT
Introduction: Gait features differ between Parkinson's disease (PD) and healthy subjects (HS). Kinematic alterations of gait include reduced gait speed, swing time, and stride length between PD patients and HS. Stride time and swing time variability are increased in PD patients with respect to HS. Additionally, dynamic parameters of asymmetry of gait are significantly different among the two groups. The aim of the present study is to evaluate which kind of gait analysis (dynamic or kinematic) is more informative to discriminate PD and HS gait features. Methods: In the present study, we analyzed gait dynamic and kinematic features of 108 PD patients and 88 HS from four cohorts of two datasets. Results: Kinematic features showed statistically significant differences among PD patients and HS for gait speed and time Up and Go test and for selected kinematic dispersion indices (standard deviation and interquartile range of swing, stance, and double support time). Dynamic features did not show any statistically significant difference between PD patients and HS. Discussion: Despite kinematics features like acceleration being directly proportional to dynamic features like ground reaction force, the results of this study showed the so-called force/rhythm dichotomy since kinematic features were more informative than dynamic ones.
Subject(s)
Parkinson Disease , Wearable Electronic Devices , Humans , Biomechanical Phenomena , Parkinson Disease/diagnosis , Postural Balance , Time and Motion Studies , Gait , BiomarkersABSTRACT
Reperfusion therapy administration timing in acute ischaemic stroke is the main determinant of patients' mortality and long-term disability. Indeed, the first hour from the stroke onset is defined the "golden hour", in which the treatment has the highest efficacy and lowest side effects. Delayed ambulance transport, inappropriate triage and difficulty in accessing CT scans lead to delayed onset to treatment time (OTT) in clinical practice. To date brain CT scan is needed to rule out intracranial haemorrhage, which is a major contraindication to thrombolytic therapy. The availability, dimension and portability make CT suitable mainly for intrahospital use, determining further delays in the therapies administration. This review aims at evaluating portable neurophysiology technologies developed with the scope of speeding up the diagnostic phase of acute stroke and, therefore, the initiation of intravenous thrombolysis. Medline databases were explored for studies concerning near infrared spectroscopy (NIRS), bioelectrical impedance spectroscopy (BIS) and Microwave imaging (MWI) as methods for stroke diagnosis. A total of 1368 articles were found, and 12 of these fit with our criteria and were included in the review. For each technology, the following parameters were evaluated: diagnostic accuracy, ability to differentiate ischaemic and haemorrhagic stroke, diagnosis time from stroke onset, portability and technology readiness level (TRL). All the described methods seem to be able to identify acute stroke even though the number of studies is very limited. Low cost and portability make them potentially usable during ambulance transport, possibly leading to a reduction of stroke OTT along with the related huge benefits in terms of patients outcome and health care costs. In addition, unlike standard imaging techniques, neurophysiological techniques could allow continuous monitoring of patients for timely intrahospital stroke diagnosis.KEY MESSAGESFirst hour from the stroke onset is defined the "golden hour", in which the treatment has the highest efficacy and lowest side effects.The delay for stroke onset to brain imaging time is one of the major reasons why only a minority of patients with acute ischaemic stroke are eligible to reperfusion therapies.Neurophysiology techniques (NIRS, BIS and MWI) could have a potential high impact in reducing the time to treatment in stroke patients.
Subject(s)
Brain Ischemia , Ischemic Stroke , Stroke , Brain Ischemia/diagnostic imaging , Electric Impedance , Fibrinolytic Agents/therapeutic use , Humans , Microwaves , Neurophysiology , Spectroscopy, Near-Infrared , Stroke/chemically induced , Stroke/diagnostic imaging , Treatment OutcomeABSTRACT
BACKGROUND: The construct of Essential Tremor plus (ET-plus) refers to patients who also have rest tremor and/or mild neurologic signs of unknown significance. It is unclear whether soft signs represent confounding factors or are useful in suspecting an alternative condition. METHODS: Using a Bayesian approach to ET-plus patients recruited in The ITAlian tremor Network (TITAN), we analyzed the probability that these patients do not have ET. RESULTS: The data of 274 ET-plus patients were extracted from the TITAN database. The majority of patients (240/274; 87.5%) had a single soft sign. The post-test probability of not having ET was different according to the specific soft sign: namely, 0.64 (rest tremor); 0.46 (questionable dystonia); 0.85 (questionable bradykinesia); 0.19 (soft gait impairment); and 0.09 (questionable cognitive issues). In patients with multiple soft signs, the post-test probability of not having ET was higher than 50% for 7 out of 11 combinations, accounting for 44.1% of subjects. Overall, the post-test probability of not having ET was higher than 50% in up to 71.5% of ET-plus patients. DISCUSSION: We have here shown that: 1) the soft signs differently contribute in modulating the probability that a patient does not have ET; and 2) the effect of multiple soft signs are not always addictive. Future studies are needed to collect prevalence figures of soft signs in different neurological disorders as well as in the elderly and to calculate their value in predicting the development of an alternative tremor syndrome.
Subject(s)
Dystonia , Essential Tremor , Humans , Aged , Essential Tremor/diagnosis , Essential Tremor/epidemiology , Tremor/diagnosis , Bayes Theorem , Cohort StudiesABSTRACT
Dystonia diagnosis is based on clinical examination performed by a neurologist with expertise in movement disorders. Clues that indicate the diagnosis of a movement disorder such as dystonia are dystonic movements, dystonic postures, and three additional physical signs (mirror dystonia, overflow dystonia, and geste antagonists/sensory tricks). Despite advances in research, there is no diagnostic test with a high level of accuracy for the dystonia diagnosis. Clinical neurophysiology and genetics might support the clinician in the diagnostic process. Neurophysiology played a role in untangling dystonia pathophysiology, demonstrating characteristic reduction in inhibition of central motor circuits and alterations in the somatosensory system. The neurophysiologic measure with the greatest evidence in identifying patients affected by dystonia is the somatosensory temporal discrimination threshold (STDT). Other parameters need further confirmations and more solid evidence to be considered as support for the dystonia diagnosis. Genetic testing should be guided by characteristics such as age at onset, body distribution, associated features, and coexistence of other movement disorders (parkinsonism, myoclonus, and other hyperkinesia). The aim of the present review is to summarize the state of the art regarding dystonia diagnosis focusing on the role of neurophysiology and genetic testing.
ABSTRACT
INTRODUCTION: The recently released classification has revised the nosology of tremor, defining essential tremor (ET) as a syndrome and fueling an enlightened debate about some newly conceptualized entities such as ET-plus. As a result, precise information of demographics, clinical features, and about the natural history of these conditions are lacking. METHODS: The ITAlian tremor Network (TITAN) is a multicenter data collection platform, the aim of which is to prospectively assess, according to a standardized protocol, the phenomenology and natural history of tremor syndromes. RESULTS: In the first year of activity, 679 patients have been recruited. The frequency of tremor syndromes varied from 32% of ET and 41% of ET-plus to less than 3% of rare forms, including focal tremors (2.30%), task-specific tremors (1.38%), isolated rest tremor (0.61%), and orthostatic tremor (0.61%). Patients with ET-plus were older and had a higher age at onset than ET, but a shorter disease duration, which might suggest that ET-plus is not a disease stage of ET. Familial aggregation of tremor and movement disorders was present in up to 60% of ET cases and in about 40% of patients with tremor combined with dystonia. The body site of tremor onset was different between tremor syndromes, with head tremor being most commonly, but not uniquely, associated with dystonia. CONCLUSIONS: The TITAN study is anticipated to provide clinically relevant prospective information about the clinical correlates of different tremor syndromes and their specific outcomes and might serve as a basis for future etiological, pathophysiological, and therapeutic research.
Subject(s)
Dystonia , Dystonic Disorders , Essential Tremor , Dystonia/complications , Humans , Italy/epidemiology , Prospective Studies , Syndrome , Tremor/complications , Tremor/diagnosis , Tremor/epidemiologyABSTRACT
Dystonia is a hyperkinetic movement disorder characterized by abnormal movement or posture caused by excessive muscle contraction. Because of its wide clinical spectrum, dystonia is often underdiagnosed or misdiagnosed. In clinical practice, dystonia could often present in association with other movement disorders. An accurate physical examination is essential to describe the correct phenomenology. To help clinicians reaching the proper diagnosis, several classifications of dystonia have been proposed. The current classification consists of axis I, clinical characteristics, and axis II, etiology. Through the application of this classification system, movement disorder specialists could attempt to correctly characterize dystonia and guide patients to the most effective treatment. The aim of this article is to describe the phenomenological spectrum of dystonia, the last approved dystonia classification, and new emerging knowledge.
ABSTRACT
INTRODUCTION: Motor complication management is one of the main unmet needs in Parkinson's disease patients. AREAS COVERED: Among the most promising emerging approaches for handling motor complications in Parkinson's disease, adaptive deep brain stimulation strategies operating in closed-loop have emerged as pivotal to deliver sustained, near-to-physiological inputs to dysfunctional basal ganglia-cortical circuits over time. Existing sensing systems that can provide feedback signals to close the loop include biochemical-, neurophysiological- or wearable-sensors. Biochemical sensing allows to directly monitor the pharmacokinetic and pharmacodynamic of antiparkinsonian drugs and metabolites. Neurophysiological sensing relies on neurotechnologies to sense cortical or subcortical brain activity and extract real-time correlates of symptom intensity or symptom control during DBS. A more direct representation of the symptom state, particularly the phenomenological differentiation and quantification of motor symptoms, can be realized via wearable sensor technology. EXPERT OPINION: Biochemical, neurophysiologic, and wearable-based biomarkers are promising technological tools that either individually or in combination could guide adaptive therapy for Parkinson's disease motor symptoms in the future.
Subject(s)
Deep Brain Stimulation , Parkinson Disease , Wearable Electronic Devices , Basal Ganglia , Humans , Parkinson Disease/drug therapyABSTRACT
Chronic pain is one of the leading causes of disability and disease burden worldwide, accounting for a prevalence between 6.9% and 10% in the general population. Pharmacotherapy alone results ineffective in about 70-60% of patients in terms of a satisfactory degree of pain relief. Focused ultrasound is a promising tool for chronic pain management, being approved for thalamotomy in chronic neuropathic pain and for bone metastases-related pain treatment. FUS is a noninvasive technique for neuromodulation and for tissue ablation that can be applied to several tissues. Transcranial FUS (tFUS) can lead to opposite biological effects, depending on stimulation parameters: from reversible neural activity facilitation or suppression (low-intensity, low-frequency ultrasound, LILFUS) to irreversible tissue ablation (high-intensity focused ultrasounds, HIFU). HIFU is approved for thalamotomy in neuropathic pain at the central nervous system level and for the treatment of facet joint osteoarthritis at the peripheral level. Potential applications include HIFU at the spinal cord level for selected cases of refractory chronic neuropathic pain, knee osteoarthritis, sacroiliac joint disease, intervertebral disc nucleolysis, phantom limb, and ablation of peripheral nerves. FUS at nonablative dosage, LILFUS, has potential reversible and tissue-selective effects. FUS applications at nonablative doses currently are at a research stage. The main potential applications include targeted drug and gene delivery through the Blood-Brain Barrier, assessment of pain thresholds and study of pain, and reversible peripheral nerve conduction block. The aim of the present review is to describe the approved and potential applications of the focused ultrasound technology in the field of chronic pain management.
ABSTRACT
The aim of this review is to summarize that most relevant technologies used to evaluate gait features and the associated algorithms that have shown promise to aid diagnosis and symptom monitoring in Parkinson's disease (PD) patients. We searched PubMed for studies published between 1 January 2005, and 30 August 2019 on gait analysis in PD. We selected studies that have either used technologies to distinguish PD patients from healthy subjects or stratified PD patients according to motor status or disease stages. Only those studies that reported at least 80% sensitivity and specificity were included. Gait analysis algorithms used for diagnosis showed a balanced accuracy range of 83.5-100%, sensitivity of 83.3-100% and specificity of 82-100%. For motor status discrimination the gait analysis algorithms showed a balanced accuracy range of 90.8-100%, sensitivity of 92.5-100% and specificity of 88-100%. Despite a large number of studies on the topic of objective gait analysis in PD, only a limited number of studies reported algorithms that were accurate enough deemed to be useful for diagnosis and symptoms monitoring. In addition, none of the reported algorithms and technologies has been validated in large scale, independent studies.
Subject(s)
Gait Analysis , Parkinson Disease , Algorithms , Gait , Humans , Parkinson Disease/diagnosis , Sensitivity and SpecificityABSTRACT
This work focuses on the design and the validation of a wearable mechatronic device for an at-home assessment of wrist stiffness in patients affected by Parkinson's Disease (PD). The device includes one actuated joint and four passive revolute joints with a high overall intrinsic backdriveability. In order to allow the user to freely move the wrist during activities of daily living, we implemented a transparent controller on the basis of the interaction force sensed by the embedded load cell. Conversely, in order to provide perturbations for estimating the wrist flexion-extension rigidity, we implemented a torque controller. Firstly, we report a pilot study that aimed at characterizing the device in terms of range of motion (ROM) allowed, transparency perceived and torque-tracking capability. Then, we present a case study in which we tested our device with seven PD patients in both drug-OFF and drug-ON conditions and we compared the measured stiffness with the one measured in fourteen healthy controls and with the outcome of the most used clinical scale (MDS-UPDRS). The device allowed to successfully estimate the stiffness as different depending on the movement direction. Indeed, extension stiffness was higher than the flexion one, accordingly to the literature. Moreover, the device allowed to discriminate both Healthy subjects from PD subjects, and PD subjects in OFF condition from PD subjects in ON condition. In conclusion, we demonstrate the feasibility of the device in measuring wrist rigidity, thus enabling the possibility to implement an at-home assessment of the PD rigidity.
Subject(s)
Parkinson Disease , Wearable Electronic Devices , Activities of Daily Living , Humans , Muscle Rigidity , Parkinson Disease/diagnosis , Pilot Projects , Wrist , Wrist JointABSTRACT
BACKGROUND: Levodopa-carbidopa intestinal infusion is an effective treatment for motor fluctuations in Parkinson's disease. However, it has been recently associated with emergent complex/atypical dyskinesias. We sought to characterize patients who developed these dyskinesias after levodopa infusion initiation, and to compare these patients to a control population with conventional motor fluctuations. METHODS: 208 Parkinson's disease patients, treated with levodopa intestinal infusion due to motor fluctuations, were screened for onset and/or worsening of dyskinesias after initiation of levodopa infusion, resistant to the routine titration, and presenting with atypical or unexpected patterns. Patients with extensive follow-up data were enrolled for a longitudinal analysis. Cases were compared to a control sample with conventional motor fluctuations in order to investigate predisposing factors, difference in dyskinesia phenotype, management strategies and dropouts. RESULTS: Thirty patients out of 208 (14.4%) reported atypical (i.e. long-lasting) biphasic, biphasic-like (i.e. continuous) or mixed (peak-dose and continuous biphasic) dyskinesias after levodopa infusion. They were compared at baseline and follow-up to a sample of 49 patients with conventional motor fluctuations on levodopa infusion. Both groups had similar demographic and clinical features, except the former having higher prevalence of biphasic dyskinesias while on oral therapy. Biphasic-like dyskinesias in nearly half the number of cases improved with increasing the dopaminergic load, while mixed dyskinesias had the worst outcome and highest dropout rate (58%). CONCLUSIONS: Atypical biphasic, biphasic-like and complex dyskinesias could hinder the course of patients treated with levodopa infusion. This study further informs the selection process of advanced therapies, particularly in dyskinetic patients.